Okay, I admit this is a WEIRD problem, but the evidence is sitting on my hard drive right now.

I was going through some previously recorded footage using the ProHD file manager (I have to use this program because I recorded in the MP4 format).

I have two shots that were filming some flowers moving in the breeze. The shots were very similar, so were ideal for the comparison I did. BOTH shots were done in 1080p25 mode, taken a minute apart. However, when playing back the second shot, I can very CLEARLY see the interlaced lines, whereas these lines are not present at all on the first shot. The difference is strikingly obvious, so I know it's not a problem relating to perception. When I hover the mouse over the clip thumbnails to check, sure enough, both shots are labelled "25p"; there is no 50i label at all.

To make sure this wasn't just a problem with the ProHD file manager program, I played both shots back in Premiere Pro at 100% scale (this is the only scale that seems to show up interlacing). And sure enough, one of the shots is interlaced; the other is not! Previewing in VLC shows exactly the same thing.

This problem also occurred on the shot taken just after the first interlaced one; but all of the shots after that, all taken within minutes of each other, are fine -- the progressive mode obviously kicked back into action (and these shots all contained sufficient movement to see if there was any interlacing).

So my question is, what on earth has triggered the JVC HM100 to revert to using the interlaced mode when I did not specify this in the menu? I know for a fact that I did not change any of these settings while recording, because the project I am doing needs to use progressive footage. And this is confirmed by the metadata which clearly says "25p".




Edit: I have just found out that while playing the footage in Premiere Pro, no matter what size I have the preview scaled to, the lines are clearly visible. This is not the case with all of the other other interlaced footage I have shot in the past with the HM100. With all other interlaced footage, the lines can only be seen at 100% scaling. The fact that I can see these lines at any scale tells me that the JVC has done something very odd -- it has recorded the interlacing as progressive footage and therefore it is as if the lines are "hard wired", and no monitor, HDTV or conversion process will be able to play back the footage so it looks acceptable!!! So along with the OIS glitch I mentioned earlier, this is ANOTHER really bad (or much worse) glitch that surfaces at random. I'm not happy right now.

Update: While I was reviewing the footage, I asked myself, "Why did I record two identical scenes with less than a minute between them?" I figured there had to be a reason for this, and the only thing I can think of is that I turned on the ND filter between both those shots. The manual says you shouldn't enable the ND filter during filming or it can cause "distortion" in the image and audio. While I already knew that and so would not have enabled it during shooting (if I did it would have shown up quite clearly), perhaps after I did enable it, I pressed the record button too quickly after that and this caused the shot to have this "distortion"? Or perhaps you shouldn't enable/disable the ND filter while the camera is on, whether shooting or not? I also just checked the audio and there is nothing out of the ordinary there. However, these theories do not explain why there was another shot after that which still shows the same "distortion" issue.

Another update: I have found this same problem again, on a totally different day. About three shots taken no more than a minute apart, in the 25p mode, have shown the interlacing issue, though after a 3-minute break, the problem goes away in the remaining shots.

I have seen that before when I was capturing using the HDMI port and recording internally. If I had the camera set to progressive it was always 1080i out. Some of the 1080/24p frames from inside the camera had interlaced lines on them but I haven't ever seen this during normal shooting and I have a few hundred hours on my camera's clock. You should try changing your shooting mode to something like 720/24p and then back to 1080/25p and see if you notice it after that. I know it has to reset some stuff in the camera every time a change like that is made and I change modes all the time.

I know that many cameras, including the JVC HM100, only output 1080i footage via the HDMI port, even if you've shot in 1080p. However this is not my problem.

I have come to learn that the way the JVC creates 1080p footage, is by somehow time-shifting every alternate line of the interlaced signal by one frame so that you end up with a "makeshift" progressive frame. This is why I can drag a 1080p clip into a 1080i timeline in Premiere Pro, and no rendering is required, because the timeline still sees the 1080p footage as interlaced, even if that is not how it looks to the eye. (Doing the reverse -- dragging a 1080i clip into a 1080p timeline -- does require rendering, however.)

Now, I imagine that this has a LOT to do with the problem I have observed. Since shooting in 720p works differently, I seriously doubt this problem would ever arise in that mode.

The question is, what has caused the camera to decide to record half of the 1080 lines one frame out-of-sync? So far I can only think of the ND filter switch somehow interfering, though that is odd since I never changed it during recording. The interlace problem seemed to go away after a few minutes of not recording anything, perhaps after I turned the camera off and turned it back on again, though I can't be sure I even turned it off during those intervening minutes.

I have also discovered that the problem can be solved by dragging the affected clip into a 1080i timeline. The interlaced lines go away after doing this. This proves that the labelled "25p" footage is in fact 50i (according to the JVC nomenclature). While it solves the problem in one sense, since an interlaced timeline with this footage would render out and play well on an HDTV, it can't ever transform the footage back into progressive unless a conversion process is used, which is undesirable.

Anthony, Can you share the footage? Just to look into it with different means.

This is a weird one. I'm going to do some tests and try to replicate.
Is there any other information you can think of that may have triggered it (besides the ND filter)?

Tim, I have been racking my brain to think of other possible things I could have done to trigger the problem. I have now found three instances where this problem has appeared, on three separate days. It has occurred in both the 1080p25 and 1080p30 modes. During or between the shots in which the problem exists, it is very possible I could have activated or deactivated the Telemacro mode (I have this set to the User 2 button). Other functions I may have activated or deactivated are the Zebra or the Focus Assist (assigned to User 3 and User 1). I just wish the JVC recorded this kind of metadata to each file to make diagnosis easier, but annoyingly it records very little such data. (My last standard definition camera, a Panasonic NV-GS400, recorded practically all settings used at the time, such as shutter speed, aperture, gain, white balance, and OIS etc. I'm surprised the JVC doesn't do this.)

Leonid, I will try to upload some samples soon. What's the best HD upload site? (I notice that the basic account for Vimeo only allows one HD upload per week, when I would like to upload two for comparison.)

Anthony,

Try RapidShare: 1-CLICK Web hosting - Easy Filehosting (http://www.rapidshare.com) - but it has a limit up to 200MByte.
You also can try Rapidshare.ru :: (http://www.rapidshare.ru) - same company but 1.5GByte limit (and faster, at least in California). I attach some brief note how to pass through Russian language menu at the end.

But first, double check - is it native MP4, I mean recorded by JVC HM100U directly to SDHC, right?

----------------
Brief instruction to upload file(s) to Rapidshare.ru :: (http://www.rapidshare.ru) (if you prefer it):

Go to Rapidshare.ru :: (http://www.rapidshare.ru), click "browse" and select file. Mark small box in next line - ("I know the rules and will follow it" - no child porn, no copyright violation, no hate crime etc - usual stuff, you are definitely safe with your clip. You may read English version in RapidShare: 1-CLICK Web hosting - Easy Filehosting (http://www.rapidshare.com), it is the same) and press button near "Browse".

After upload, record the "URL" line - it has a direct reference to your file. During download it force you wait 60secs and suggest download after that (like RapidShare: 1-CLICK Web hosting - Easy Filehosting (http://www.rapidshare.com) does).

Thanks for that, Leonid.

You can download an 18 MB sample of this video at

http://dl4.rapidshare.ru/1123902/29914/720p25.m2v

Please note that I downconverted the video to 720p25, but despite this, the end result is that it looks EXACTLY the same as the 1080p25 version (the lines are clearly visible). I played it using VLC. It is an MPEG-2 file.

This was 25 p and a still frame taken from the footage you can clearly see lines

Please note that I downconverted the video to 720p25

Anthony, is it possible to see an original clip from camcoder? The down converting process may destroy some valuable information or even add something else.

I will look your file later.

I just had the same thing happen to me the other night. It was a band performance in a bar and I thought it might have something to do the shutter interaction with the neon light that was in the window and sending some light to the scene, although it didn't make too much sense.

Unique camera parameters about this shoot:

The lighting there was abysmal, so I had the camcorder on high gain (9 or 18 db) , and in 1080 30P aperture wide open with a 1/30th shutter speed. I usually tape in 1/60 second shutter speed or faster. On other thing unique about this shoot is that I had the camcorder plugged into the AC/DC power supply, so I wouldn't have to worry about changing batteries during the 3.5 hour shoot.

This glitch remained on about 3 hours of HM100 video I shot that night over the span of about 15 files, it didn't come and go as others have mentioned. (used file/card spanning, I turned off record only a few times during breaks. The spanning didn't work quite as reliably as expected, maybe I'll address that into a different forum post.

The recording was not critical, or I would have cared a bit more and tried to correct the problem in post, but as it was it was just for feedback for the band and for a SD DVD to give the band.

Looking at this forum thread it makes sense that there is some bug that is causing this. I did a more important shoot the following morning and it seems everything was recorded normally (I had it mostly in "Auto").

Does it have to do with gain, shutter speed, AC power, or is it some random thing? Also the problem is you really can't tell anything is wrong with the LCD. I looked at the original clips directly from the HM100 through an HDMI port and I can't see interlacing like I can when playing the files via FCP or Quicktime. Maybe the problem a glitch in the file information that the HM100 writes and that QT uses, I don't know. All I know is that playing clips that appear to be recording in the same format, on my Mac with 10.5.7 this particular night's 1080 30P files had this problem.

I tried seeing if putting the clip into a XDCAM EX 60i timeline to see if I could fix it, and it didn't for me. It looks like the interlacing is embedded into progressive video, there's no way to 'deinterlace it.' so the footage is junk. It seems that the problem is not just that you have interlaced instead of progressive, it's that you have interlaced footage put into progressive frames. You can't fix it in post. If you had an important shoot with a lot of movement, you'd be screwed.

I'm just hoping this is an isolated incident. I haven't closely looked at all the previous footage I've shot with the HM100. I think most of it is OK though, but I'm not certain.

Attached is a screen capture of an area of an arm moving which clearly shows the interlacing (this was from my FCP canvas window, and not the whole frame)

So Tim, JVC or others, if you are listening, please, please investigate. If you want I can upload some files to one of your FTP sites so you can investigate further.

Keith,

Is it MOV?

Can you share the original clip if it is not MOV file? (I can't work with .MOV files, all my stuff is for MPEG/MP4).

P.S. As for now I see the common - low light (ND filter in Anthony case). It seems that under shutter speed 1/30 there is some jadder...

Leonid, it is a .MOV, I shoot in that mode because I use a Mac. I do remember that when I first turned the HM100 on I believe I had the ND on. I don't think I had it recording right at that point, but I did turn the ND off while it was still powered on. This shouldn't be an issue though, I do it all the time.

Probably we need to do more tests to see if we can repro it. Problem it is doesn't seem to show up easily unless you copy over the MOV /.MP4 files.

I want to try some tests. But first, I need to understand a few things.

1. Are we testing for 24p, 25p, and 30p in 1080 mode?

2. Does the HM100 capture 24p natively (24 frames, not 24p over 60i)

3. Does the HM100 capture 30p natively (30 frames, not 30p over 60i)

I can not get a clear answer from the documentation on how the HM100 handles these.

Thanks.

1. Are we testing for 24p, 25p, and 30p in 1080 mode?
It seems like his were 25p I have seen my own under 24 but I shoot a lot under that and only noticed it when and HDMI cable was also plugged in. I am talking about the 24p the camera recorded, not what came out of the cable.

2. Does the HM100 capture 24p natively (24 frames, not 24p over 60i)
Native
3. Does the HM100 capture 30p natively (30 frames, not 30p over 60i)
Native

To Elvis and Andy:

If set to progressive, the HM100 should be recording native progressive and saving the data in the XDCAM EX format.

The XDCAM EX codec progressive mode is true progressive, it's not an interlaced format that combines or skips frames like some of the tape-based DV and HDV codecs which was an artifact and limitation of tape transport recording.

I personally use 1080 30P and this is where I experienced my problem, which is that it seems there is interlacing artifacts embedded into file that should be 'progressive.' Whether this is happening in the camcorder and is part of the actual video data or it's incorrectly recording interlaced and marking it as progressive, something is going wrong at the camcorder end.

Anthony,

I looked into your clip. It looks like it is not a pure interlaced issue. The grass blades in front plane out of focus have clear light-green/dark-green long lines which lasts a couple of frames. But it should not be at all because grass is out-of-focus. Moreover, lines with interlacing problem should exist only 1 or 2 frames at the same place but not more.

It is more interesting look into original clip because this one has a trace of some re-encoding, it has Format profile: Main@High-1440 which is not available for MP4 shooting. That may underline the problem.

Anthony,

I looked into your clip. It looks like it is not a pure interlaced issue. The grass blades in front plane out of focus have clear light-green/dark-green long lines which lasts a couple of frames. But it should not be at all because grass is out-of-focus. Moreover, lines with interlacing problem should exist only 1 or 2 frames at the same place but not more.

It is more interesting look into original clip because this one has a trace of some re-encoding, it has Format profile: Main@High-1440 which is not available for MP4 shooting. That may underline the problem.

Leonid, please understand (and trust me) that if you had watched that clip as 1080p25, it would have looked IDENTICAL. I don't want to upload the whole, original file because then it would be about 200 MB in size, and I don't have a fast internet connection. The next best thing I can do is upload a shorter MPEG-2 file in 1080p25 mode. I have played the same original clip in the JVC ProHD Clip Manager and it looks exactly as the clip you now see. I will upload that version soon.

Also, about the file format being "Main@High-1440" -- I just realised that that was the "Level" that was set as the default in Adobe Media Encoder. I could have just chosen "High".


I tried seeing if putting the clip into a XDCAM EX 60i timeline to see if I could fix it, and it didn't for me. It looks like the interlacing is embedded into progressive video, there's no way to 'deinterlace it.' so the footage is junk. It seems that the problem is not just that you have interlaced instead of progressive, it's that you have interlaced footage put into progressive frames. You can't fix it in post. If you had an important shoot with a lot of movement, you'd be screwed.

Yes this is what I was trying to convey in my original post about the interlacing being essentially "hard wired" into the progressive frames.

Interesting that FCP couldn't solve the problem by placing it into a 1080i timeline. Premiere Pro doesn't have a problem fixing it for me. Maybe you should try Premiere? I know Premier can handle .mov files. (I also use a Mac, btw, but prefer Premiere since it integrates so well with After Effects and Encore, which I use frequently.)

The problems I've had have all happened using battery power. It has only remained for two or three successive clips, then goes away on its own, even if the first clip in which the problem no longer exists is only a matter of a minute or two after the previous problem clip. I can't think what I've done between those clips to correct the problem. I doubt I have turned the camera off in that intervening time. From now on, when recording certain clips, I'm going to announce to the camera the main settings I have and what I've changed. That way when the problem inevitably arises again, I can see if there's a pattern.


1. Are we testing for 24p, 25p, and 30p in 1080 mode?
It seems like his were 25p I have seen my own under 24 but I shoot a lot under that and only noticed it when and HDMI cable was also plugged in. I am talking about the 24p the camera recorded, not what came out of the cable.

2. Does the HM100 capture 24p natively (24 frames, not 24p over 60i)
Native
3. Does the HM100 capture 30p natively (30 frames, not 30p over 60i)
Native

I thought the JVC captured 30p over 60i. This would explain why I can drop a 1080p30 clip into a 1080i60 timeline in Premiere Pro and no rendering is required, because Premiere can "see" the separate fields.

This is a fascinating thread, has anyone contacted JVC about this, maybe they know exactly what is going on, who knows, maybe this is something slated already to be fixed in the firmware update? But it sure seems like someone should contact them, as the camera certainly shouldn't be operating in this way, right?

Tim

I called today so they now know about the issue .So I will hopefully hear from their engineers
Doug

Leonid, please understand (and trust me) that if you had watched that clip as 1080p25, it would have looked IDENTICAL.

Anthony, it is not a matter of trust, I believe you have seen something wrong. It is all about localizing the problem and find the way around it.

What puzzles me - there are 3 unexplained issue to mark the problem as "interlaced".

1) Long zigzag lines in front plane grass blades. In frame-by-frame advance it seems that blades moves slowly and it's edges are out-of-focus. However, zigzag lines are long and clear.

2) Zigzag lines and spaces between them consist from TWO pixels width. But fields actually interleave each other pixel line, that is a definition. I double check with my 1920x1080 clips and yes, zigzag lines in progressive still shot (postprocess from interleaved i60) are only one pixel width.

3) You said you downconverted a 1080 clip. But in most of downconverting procedures the zigzag lines will be blurred away because of a resolution drop (from 1080 to 720 - it is a 50%). In my knowledge the only downconverting procedure which could keep interleaved zigzags is the downconverting from interleaved stream and converting it to progressive. In some ways it may downgrade separately fields and turn to progressive later.

Now, because you are shooting in MP4 and I can't successfully analyze Keith's MOV files I have some interest in your original clip - that is a reason. I browsed through my shots (1+ month) but I don't see this problem. However, my experience with p30/p25 is limited because I shoot in i60 and use 720p60 for room/evening shots.

Leonid, would a 1080p25 clip in MPG format suffice? Adobe Media Encoder only allows me to output such a clip in the MPEG-2 format.

On the other hand, I have a 53 MB clip which is 12 seconds long that has this interlacing issue (or whatever you prefer to call it). I could upload this clip, completely unaltered, for you to analyse. But be warned that there is not a lot of movement in this clip so it is harder to see the problem than in the clip you've already seen.

I have a 53 MB clip which is 12 seconds long that has this interlacing issue (or whatever you prefer to call it). I could upload this clip, completely unaltered, for you to analyse. But be warned that there is not a lot of movement in this clip so it is harder to see the problem than in the clip you've already seen.

That would be great! Just point us to approximate time of trouble.

Leonid, here is the link of the unchanged MP4 file: http://dl4.rapidshare.ru/1126863/66133/000_0063_01.MP4

It is only 12 seconds long. Like I said before, the artefacts are not as easy to see as in the previous clip, though they are definitely there. Also the clip is pretty noisy because I think I forgot to turn the ND filter off.

Leonid, here is the link of the unchanged MP4 file: http://dl4.rapidshare.ru/1126863/66133/000_0063_01.MP4

It is only 12 seconds long. Like I said before, the artefacts are not as easy to see as in the previous clip, though they are definitely there. Also the clip is pretty noisy because I think I forgot to turn the ND filter off.

Anthony,
I can confirm the interlaced image issue. Like you, that's what i'm calling it – as the evidence from my camera is very compelling. I looked at your footage and could see the same thing I have seen with my HM100. It's hard to see in your footage as there's little movement in your shot but it is there nonetheless.

I have had it in all files of 2 out of 4 shooting sessions that I have used the camera when I had it set to 1080 30p.

See my full size frame grab exported directly from a file opened in QT Player. No processing of any kind apart from jpg-ing it for upload. You can see the pronounced lines which are one pixel high:

http://www.flickr.com/photos/zenlessness/3796477140/sizes/o/

JVC have my camera for this and other reasons and I hope to hear from them about the interlacing problem any day.

Does this happen in only 30p, or in 60p and 24p, too.

Please let us know what JVC says and if we need press for a recall.

There is something strange about. And suddenly word is silent on the training DVD that is past due coming out (with a single copy turning up on Australian Ebay).

It will be interesting to see what the announced firmware fix will be at the end of the month... or if news on that just goes silent, too.

Does this happen in only 30p, or in 60p and 24p, too.

.

I've only really tested the camera in 30p. The camera does 60i not p.

I'll keep you posted.

The camera does 60i not p.

The camera DOES do 720P60...

So Tim, JVC or others, if you are listening, please, please investigate. If you want I can upload some files to one of your FTP sites so you can investigate further.
I've been trying to replicate the issue but haven't been able to. Did anyone else stumble upon a sure-fire way to do it?
There is something strange about. And suddenly word is silent on the training DVD that is past due coming out (with a single copy turning up on Australian Ebay)
The delay on the release of the DVD has nothing to do with this. The DVD's were replicated over a month ago. The delay was simply in getting the free tutorials uploaded and the sales site built and implemented, which went live late yesterday. http://www.dvinfo.net/prohd You will start to see the DVDs pop up on ebay and on dealer websites this week. There's no conspiracy with JVC and I'm sure they have likely read this thread and are looking into it themselves.

The delay on the release of the DVD has nothing to do with this. The DVD's were replicated over a month ago. The delay was simply in getting the free tutorials uploaded and the sales site built and implemented, which went live late yesterday. http://www.dvinfo.net/prohd You will start to see the DVDs pop up on ebay and on dealer websites this week. There's no conspiracy with JVC and I'm sure they have likely read this thread and are looking into it themselves.
Thank you for the update. I have ordered a DVD!

The camera DOES do 720P60...

My mistake – you are quite correct.

With regards reproducibility, I have gone through my footage again. I thought that it might be to do with changing the capture system from one mode to another but the evidence shows otherwise as I have only changed twice to 720 30p and then back to 1080 30p.

The interesting thing is the way that camera produces interlacing in one shot taken 1 minute after a normal progressive scan shot.


I use caps only to clearly mark the interlaced shots below:

1st batch of 4 shots - no interlacing . 1080 30p

2nd batch of 15 shots. First eight no interlacing and then on the next shot (with 1 minute gap between) I get INTERLACING which continues for the remaining 6 shots. All 1080 30p

3rd batch. 23 shots all progressive scan.

4th batch. 13 shots 1080 30p. the last shot has INTERLACING.

5th batch . A bunch of 720 30p with no interlacing

6th batch. 15 shots 1080 30p all INTERLACED.


Also I can confirm that QT Player shows that it considers the interlaced footage to be 1080 30p in Movie Inspector.

I can see no pattern here.

I will relay any information I get from from JVC.

New firmware has been posted on the JVC site. Please download and update your camera, and advise if the problem continues to exist.

JVC Professional Customer Support - Product Upgrades (http://pro.jvc.com/prof/support/productupdate.jsp)

Dave Walton
Asst. Vice President Marketing Communications
JVC U.S.A.

Dave, many thanks for the firmware info. JVC have looked at sample footage and straight away said that they'd replace the camera. Thanks JVC – for the good service and customer support.

Weird, but a happy ending. (we hope) .

My experience their JVC's Pro-products customer service is very good. I emailed JVC once at midnight about a question and I got a phone call the next day from from their US lead tech support guy who left a message answering my question on my answering machine with his cell phone number if I had a follow up question. So we hope it's just some weird thing that was defective and that your replacement will cure your ills. But yeah.. it does sound bizzare. almost like a bug in the programming that the firmware update might have taken care of.

Simon, it's good that JVC replaced your camera! I have only just got the firmware update, but if I notice the problem continuing, I'm going to try and get a replacement also.

Btw, for those who asked, this issue, as I stated much earlier, was present in BOTH the 1080p25 and 1080p30 modes.

Hello
I had this problem also in 25P but only when shutter was in 25.

I want to understand, does they replace the camera if they see the problem?

thanks

Amos

Amos,

In my case the camera was with JVC and I sent an example file from the camera that displayed the problem. Better is to be able to give them the SD CARD with the original file(s) on as it also contains (xml data) files that have the clip settings info. And, in answer to your question, in my case they replaced the camera.

OK, I just got the problem again, for the first time in a while. It happened with I did a custom white balance (pushing the button on the front of the cam) just prior to shooting. I think the last time it happened I did the same thing. Other times I've just prior presets, for the most part.

The original SDHC card has long been reused so can't access that anymore.

Unfortunately this time it was on a less trivial gig. The only saving grace is that it was my B-CAM, my Sony EX1 being the A Cam. I'm seeing if I can load the clips back onto a SDHC card and see if the camera outputs the files over component correctly.

Anyway I'm going to update the firmware but since you don't know if it happened until you get it transferred from your computer there is no way to know when it's happened until it's too late. Not the best problem to have.

Has anybody else experienced this problem after the firmware update?

Keith, I haven't used my JVC a lot since installing the update, so I can't comment just yet as to whether it addresses the problem. However I will very soon be going through footage which was shot after the firmware update, so I should have a pretty good idea of whether it solves the problem or not.

OK, I just got the problem again, for the first time in a while. It happened with I did a custom white balance (pushing the button on the front of the cam) just prior to shooting. I think the last time it happened I did the same thing. Other times I've just prior presets, for the most part.

Keith, have you considered sending the camera and the clip back to JVC? As you may be aware from my experience they replaced the camera.

Hi Simon

I haven't, but I may. I sent a personal message to David Walton of JVC about what to do and he hasn't responded.

I may contact JVC, not sure exactly who to contact though in the U.S.

How long did it take JVC to get you a new one? And how long had you owned the HM100 before returning it?

Thanks.

-Keith

Hi Keith,

I had only had the camera two weeks before I returned it to the shop. It took another 10 days for JVC to respond on other matters and then I pressed them (through the shop) on the interlacing issue. Once I had emailed an example clip to JVC UK the head of department agreed to replace it straight off. It then took maybe two to three days for the new camera to get to the shop for pick up.

I haven't been able to use it much since but a friend used it for two weeks on a shoot and I didn't see any interlacing issues on the footage he showed me. Needless to say that if I see interlacing again the camera will go back again.

Well I can now conclude that the firmware update has not solved the problem, or maybe it "half" solves it, by making the problem less common.

I have been through hundreds of more recent clips that were shot after the firmware update, and I have encountered my first interlaced clip even though it was shot in 25p mode. Having said that, the problem does seem to be more rare now, though that could just be coincidence.

Anthony,

What are you going to do about it? Are you going to try to get a replacement HM100 from JVC? I haven't looked over every single clip I've done with the HM100, but I'm hoping I'm 'lucky' like you.

So far my track record though isn't good though. Of dozens of sessions I've done, 2 have turned out with this problem, which are ones I've had to use in edits, one paying, one not.

That being said, if even 1 important clip has this problem, it's a pain to resolve it, though possible. I think I may have to get a AJA or Blackmagic HDMI input card to transfer these problem clips when they occur, and probably everybody shooting in Progressive should do the same. Maybe JVC should include one with every HM100?

I shot some stuff yesterday, a little over an hour, fairly important stuff, and the interlaced problem was on about 3/4 of it. These were very high gain manual aperture. I'm thinking that's the trigger. The previous situations I've been in I believe I had adjusted the gain higher than 0 and was in manual exposure mode. I don't think I've ever experienced the problem in automatic exposure mode.

Those out there that had the problem and got your HM100 replaced, have you experienced it again? I'm about to send it back for a replacement. I'm pretty annoyed.

I wanted to recount my experience in getting this problem fixed by JVC. After the fix I don't seem to be experiencing the problem, whereas prior to the fix I was experiencing it all the time.

Craig Yanagi of JVC (he's posted here before) contacted my via email at the end of November expressing his concern for the problem I was having. After going back and forth with him, supplying sample clips, meta info, etc, to him to his technical people, JVC determined that the problem was indeed an inherent problem with the HM100, and that they were actively working on a solution.

After about a month or so Craig told me a fix was ready, and it was a firmware fix that required that my HM100 be sent into the Southern California service center (Cypress) and would take a few days to fix. Apparently this firmware update is not user-installable. Craig was even generous enough to offer a loaner HM100 for me to use while mine was in the shop (unfortunately I experienced the problem even with his loaner unit, but hey, at least I had B camcorder).

The JVC service center fixed it in a couple of days, and while they were at it they looked at my loose LCD hinge and fixed that as well as it was in warranty. They even just sent me a new viewfinder lens protector to replace my scratched one (it was backordered and they just received it.)

Overall, though the initial problem was pretty annoying, and I wasted a lot of time dealing with post-processing the faulty interlaced video to get it to be progressive, I have to say that my experience with Craig and JVC has bee exemplary. Though I have some criticism of the design of the HM100, I cannot fault Craig Yanagi and JVC in owning up to the problem and doing everything they could as fast as they could to address the problem.

I can truthfully say that my customer service experience with JVC regarding this issue rivals the best experiences I've had with any tech company, and Craig and JVC need to be commended for taking care of me and this particular issue with superb attention and professionalism.

I asked Craig if he thought it was ok for me to post my experience, and he said that I should. Hopefully this won't result in a flood of service firmware updates to JVC, but I think that if you think you've experiencing the interlacing problem, you should considering getting the update.

I also want to thank Chris Hurd and DVInfo for providing this forum, I think the posts here resulted in me and my issue getting attention, and I'm grateful to be a part of the DVinfo community.

-Keith Moreau
SilverSpot Media
silverspotmedia.com

Damn,...I was completely unaware of this thread when I noticed mine having the same problem the other night. (I started a new thread and was told to search to find this one) My cam is was purchased in June, and is running the latest SD card class 10 upgraded firmware,..so I guess I gotta call JVC on Monday.

I noticed the problem is shutter speed dependent. For the interlacing to happen I need to shoot 30P with a shutter set at exactly 1/30th of a second. Any other speed with force the proper progressive sensor readout.

Damn, (again)

CT

It's important to note that the "firmware upgrade (Aug 2009)" that is available for download from JVC will not fix this problem.

I spoke with a JVC Canada rep this morning about the issue and was told to ship it to Toronto but he warned me that the upgrade would take more than a week because they have to order the replacement motherboard from Japan.

In Keith Moreau's post above he states, "apparently this firmware update is not user-installable." Thats' because in this case JVC uses "firmware upgrade" to literally mean they have to replace the hardware.

If you own the HM100 it would be wise to test for this behaviour before your warrantee expires. The fix after that time will likely be very expensive. Especially important if you intend to sell the camcorder in the future.

I feel badly for JVC because otherwise their products and service have been great in my experience. This will be a very costly warrantee issue for them.

On the other hand, I bought my HM100 after the problem was acknowledged and they probably should have recalled them immediately for testing and replacement. As it is, I will be without my primary camera and will be forced to rent if necessary, after having spent the money to avoid doing so.

At this point it seems that JVC is placing the onus on the owners to come forward in reporting that their camcorder also has the problem. I don't mean to bad-mouth JVC but it is what it is.

I'm just glad that I didn't have any customer footage or unique events shot at the setting that reproduces the problem, because I easily could have. Whew!

Thank you Jordan for posting this!

I believe this is an incredibly important issue that all GY HM-100 owners need to watch out for.

I also feel bad for JVC about this. However, I AM very glad that they acknowledge the problem and are fixing under warranty. Bravo JVC!

Mine is going back tomorrow.

CT

I have this problem as well, unfortunately discovered this on a paid gig. I usually shoot 60i but switched to 30p and 1/30 to match a locked off EX1, it was a musical performance so lighting was pretty low.

When I got home and looked at my footage I could not believe what I was seeing - anytime there's movement in the frame jagged lines appear. Footage is pretty much unusable and obviously the performance (2 hours) cannot be re-shot.