Received the Nano/Testing/Rocks

Dan Keaton · August 16th, 2009, 07:04 AM

Quote:

Originally Posted by Rafael Amador

Hi Dan,
i know that you can record with the data-rate that you want (from 19 to 220 Mbps),
I'm talking about exporting from FC with the XDCAM 422 50Mbps.
Cheers,
rafael

Dear Rafael,

If you have the appropriate HD-SDI outputs on your computer, you can play your timeline in Final Cut Pro out to HD-SDI and have the nanoFlash record the uncompressed sequence in any of our formats/mode/options.

This method allows you to keep all of the quality.

Dan Keaton · August 16th, 2009, 07:13 AM

[QUOTE=Barlow Elton;1226394]
Believe me, I am aware that rendering causes the bit rate to conform to whatever the codec dictates in a FCP timeline,

however, what I've found is that unrendered high bit rate nanoFlash files do keep their original quality, even when editing in a XDCAM 422 50mbs sequence.

And I don't mean that in the sense that they look fine, but that they actually export as self-contained QT movies with the 100mbs rate....

Dear Barlow,

This is a great discovery and excellent news.

Personally, I was not aware of this.

Thank you very much for reporting this!

This discovery, combined with the other option of sending out your full uncompressed rendered timeline through HD-SDI (or possibly HDMI) to the nanoFlash, gives you some powerful options.

Johan Hough · August 19th, 2009, 12:01 AM

I recently started recording wildlife, mostly Fynbos flowers and the mountains where it is found. I use a Sony PMW-EX1 which is a fantastic camera for this purpose.

However, having heard of the Nano Flash recorder, I was wondering what improvements in picture quality I could expect in the work I'm doing.

I'm aware of the advantages in recording fast, demanding scenes and in keying. Would I see a general improvement, even in less demanding scenes? Also, would there be a visible improvement in a static wide shot with lots of detail?

Thanks!

Dan Keaton · August 19th, 2009, 06:08 AM

Dear Johan,

The answer is yes. But I think it is best to have some of our users answer your question.

From a technical point of view, your native Sony EX1/EX3 footage is 4:2:0.

The nanoFlash records in 4:2:2 and this provides more color information.

The additional color information will be in your recorded images.

I hope others will share their experiences.

John Richard · August 19th, 2009, 08:14 AM

We employ a Canon XLH1 and XHG1 and the side by side comparisons between the 25mbps 4:2:0 HDV footage and and the 4:2:2 50/100mbps FlashXDR and Nano footage is huge in every instance. The color is just so much richer and the absence of macro-blocking and diagonal line "jaggies" in the footage out of the Convergent-Design recorders is just wonderful. It puts your work in a whole different class.

And often overlooked is the boost to audio from 16bit to 24bit which holds up better to manipulation in post as well as allowing you to protect yourself better during recording by going for a little lower average level giving you extra headroom against sudden amplitude changes and clipping. And instead of having to use separate devices like a Fostex FR2 or the like to record double for the better sound values, you get your 24 bit audio cake and eat it too with these recorders.

And as for post - capturing tape - for get about it! huge time savings - like moving files from one hard drive to another instead of running tape thru a deck at 1 times real time.

And the cost of the recording media per GB is also a big win for the C-D recorders.

I would guarantee you that the comparison between 4:2:0 35mbps and the various 4:2:2 higher data rates out of the XDR and Nano is a different world.

Spending about $3,000 takes our sub $10k cameras into the big daddy arena and beyond and goes a long way of making up for the lense and sensor size. There are folks using these devices with Vipers!!! that should tell us something.

Aaron Newsome · August 19th, 2009, 12:18 PM

Quote:

Originally Posted by John Richard

There are folks using these devices with Vipers!!! that should tell us something.

Yes there are. I've done my research and there is nothing, absolutely nothing on the market even close to the Convert Designs in terms of price / performance.

Battery powered, portable, HD-SDI, full raster 1920x1080, recording to inexpensive media is unheard of at this price.

With the instant upgrade and improvement to ANY camera, these things should be flying off the shelf.

I used to look at my HDV footage and think, "that doesn't look so bad". Compared to full raster, high bitrate files coming from the XDR, there is no comparison.

I shot an event last Thursday with 2 cameras. One camera was an HDV camera (recording to a Firestore in native HDV), stuck to some truss with a bogen super clamp. The other camera I handheld for the entire event and that was recording to a Convergent Designs box
My idea was that I would get some decent cutaways from the truss clamped camera. After laying both footage into a multiclip timeline in FCP, I can't bring myself to intercut the footage. The difference in quality is just too dramatic. The lack of detail and jaggies in the HDV footage is noticeable, especially when compared to the 100Mbps, 4:2:2, full raster footage.

I may have to get another CD box or sell my HDV camera.

David Issko · August 19th, 2009, 02:37 PM

Quote:

Originally Posted by Aaron Newsome

I may have to get another CD box or sell my HDV camera.

This is exactly what is going through my mind with my Sony V1 camera! (compared to my EX3 with my very soon to be shipped nanoflash)

Rafael Amador · August 20th, 2009, 07:02 AM

For my self is not a question about if there is something that can record with higher quality than the NANO or not.
The point is that (unless you intend to go to the big screen) more quality that the one offered by the NANO, is unnecessary.
TV, Video and the electronic components (Luminance and Chrominance) are based in the capabilities and shortcoming of the human eye.
Why to record more quality that the one your eyes can appreciate?
Why to manage more information than the one your displays can shows?
To get your Y' recorded in 10b sure would be great, but you won't see much difference with the color components recorded in 8 or 10b.
I work in video since 1.983 and, except the Ampex Quad, I've worked with all the mayor formats in the market. From my experience I have to consider the NANO like a dream come true.
Best,
rafael

Aaron Newsome · August 20th, 2009, 02:36 PM

Quote:

Originally Posted by Rafael Amador

The point is that (unless you intend to go to the big screen) more quality that the one offered by the NANO, is unnecessary.
TV, Video and the electronic components (Luminance and Chrominance) are based in the capabilities and shortcoming of the human eye.
Why to record more quality that the one your eyes can appreciate?
Why to manage more information than the one your displays can shows?

I've tried keying 4:1:0 and 4:2:0 subsampled footage, and honestly, my experience with doing it makes me want disagree with you. While I was able to get a reasonably workable key most of the time, it really is a lot more work than if I had all the color info in the file.

There are people who will tell you that you can key DV and HDV footage with no problems. I just tend to think it's harder to do with most of your color information being thrown away right at the capture stage.

Just another reason I'm going to love keying with the convergent designs box.

Wait, I just read your post again. I think we actually agree.

Rafael Amador · August 20th, 2009, 08:34 PM

Hi Aaron,
for acquisition, a kind of hybrid 422 codec: 10b Y' and 8b C'b/C'r would be perfect.
Cheers,
Rafael

Barlow Elton · August 20th, 2009, 08:52 PM

[QUOTE=Dan Keaton;1229395]

Quote:

Originally Posted by Barlow Elton

Believe me, I am aware that rendering causes the bit rate to conform to whatever the codec dictates in a FCP timeline,

however, what I've found is that unrendered high bit rate nanoFlash files do keep their original quality, even when editing in a XDCAM 422 50mbs sequence.

And I don't mean that in the sense that they look fine, but that they actually export as self-contained QT movies with the 100mbs rate....

Dear Barlow,

This is a great discovery and excellent news.

Personally, I was not aware of this.

Thank you very much for reporting this!

This discovery, combined with the other option of sending out your full uncompressed rendered timeline through HD-SDI (or possibly HDMI) to the nanoFlash, gives you some powerful options.

An update to this subject:

What I've found is that indeed you can drop a 100 or even 160mbs long GOP nano/XDR QT clip into FCP and if you let the NLE auto-conform your settings to the codec *it thinks you're working with*, (wink wink) in this case XDCAM 50mbs 1080i 60, it will allow you to export without reconforming the GOP structure of the clip (i.e. recompressing) with the exception of the beginning 15 or 30 frames of an edit/cut. (assuming you did no other image manipulation except a cut)

Sorry, that's a long-winded way of saying it won't recompress the clip unless you're altering the image (CC, filtering, transitions, etc) or you have an edit point on the timeline, in which case it has to reconstruct the GOP at that edit point.

I did some extensive testing with one of my codec torture test clips and you could clearly see where the image had been recompressed to 50mbs XDCAM and where the original (uncompressed quality) clip was left alone.

Barlow

Aaron Newsome · August 21st, 2009, 07:46 AM

This is why my workflow includes getting the rough cut on the timeline in FCP and immediately using media manager to create new clips for me, in a new project. Either 10 bit Uncompressed or 10 Bit ProRes422 HQ. Once I have the new project, I'm confident that the quality will remain through compositing, color grading, etc.

You need to make sure you check "Delete Unused Media" in media manager though, otherwise it will export the complete length of your original clips.

Barlow Elton · August 23rd, 2009, 02:58 PM

Yup, I think rough cutting in an XDCAM timeline, and then using the media manager to transcode to ProRes for final edits/corrections/outputs, etc, is a great workflow. Much easier than converting everything to ProRes from the beginning. I think uncompressed 10bit is overkill for most anything though, especially if it's a lengthy project.

August 16th, 2009, 07:13 AM	#32
Dan Keaton Inner Circle Join Date: Dec 2002 Location: Augusta Georgia Posts: 5,421	[QUOTE=Barlow Elton;1226394] Believe me, I am aware that rendering causes the bit rate to conform to whatever the codec dictates in a FCP timeline, however, what I've found is that unrendered high bit rate nanoFlash files do keep their original quality, even when editing in a XDCAM 422 50mbs sequence. And I don't mean that in the sense that they look fine, but that they actually export as self-contained QT movies with the 100mbs rate.... Dear Barlow, This is a great discovery and excellent news. Personally, I was not aware of this. Thank you very much for reporting this! This discovery, combined with the other option of sending out your full uncompressed rendered timeline through HD-SDI (or possibly HDMI) to the nanoFlash, gives you some powerful options. __________________ Dan Keaton Augusta Georgia

August 19th, 2009, 06:08 AM	#34
Dan Keaton Inner Circle Join Date: Dec 2002 Location: Augusta Georgia Posts: 5,421	Dear Johan, The answer is yes. But I think it is best to have some of our users answer your question. From a technical point of view, your native Sony EX1/EX3 footage is 4:2:0. The nanoFlash records in 4:2:2 and this provides more color information. The additional color information will be in your recorded images. I hope others will share their experiences. __________________ Dan Keaton Augusta Georgia

August 19th, 2009, 12:01 AM	#33
Johan Hough Tourist Join Date: Aug 2009 Location: Cape Town Posts: 1	I recently started recording wildlife, mostly Fynbos flowers and the mountains where it is found. I use a Sony PMW-EX1 which is a fantastic camera for this purpose. However, having heard of the Nano Flash recorder, I was wondering what improvements in picture quality I could expect in the work I'm doing. I'm aware of the advantages in recording fast, demanding scenes and in keying. Would I see a general improvement, even in less demanding scenes? Also, would there be a visible improvement in a static wide shot with lots of detail? Thanks!

August 19th, 2009, 08:14 AM	#35
John Richard Major Player Join Date: May 2006 Location: Incline Village, Nevada Posts: 604	We employ a Canon XLH1 and XHG1 and the side by side comparisons between the 25mbps 4:2:0 HDV footage and and the 4:2:2 50/100mbps FlashXDR and Nano footage is huge in every instance. The color is just so much richer and the absence of macro-blocking and diagonal line "jaggies" in the footage out of the Convergent-Design recorders is just wonderful. It puts your work in a whole different class. And often overlooked is the boost to audio from 16bit to 24bit which holds up better to manipulation in post as well as allowing you to protect yourself better during recording by going for a little lower average level giving you extra headroom against sudden amplitude changes and clipping. And instead of having to use separate devices like a Fostex FR2 or the like to record double for the better sound values, you get your 24 bit audio cake and eat it too with these recorders. And as for post - capturing tape - for get about it! huge time savings - like moving files from one hard drive to another instead of running tape thru a deck at 1 times real time. And the cost of the recording media per GB is also a big win for the C-D recorders. I would guarantee you that the comparison between 4:2:0 35mbps and the various 4:2:2 higher data rates out of the XDR and Nano is a different world. Spending about $3,000 takes our sub $10k cameras into the big daddy arena and beyond and goes a long way of making up for the lense and sensor size. There are folks using these devices with Vipers!!! that should tell us something.

August 20th, 2009, 07:02 AM	#38
Rafael Amador Major Player Join Date: Jun 2009 Location: Vientiane (Lao PDR) Posts: 349	For my self is not a question about if there is something that can record with higher quality than the NANO or not. The point is that (unless you intend to go to the big screen) more quality that the one offered by the NANO, is unnecessary. TV, Video and the electronic components (Luminance and Chrominance) are based in the capabilities and shortcoming of the human eye. Why to record more quality that the one your eyes can appreciate? Why to manage more information than the one your displays can shows? To get your Y' recorded in 10b sure would be great, but you won't see much difference with the color components recorded in 8 or 10b. I work in video since 1.983 and, except the Ampex Quad, I've worked with all the mayor formats in the market. From my experience I have to consider the NANO like a dream come true. Best, rafael

August 20th, 2009, 08:34 PM	#40
Rafael Amador Major Player Join Date: Jun 2009 Location: Vientiane (Lao PDR) Posts: 349	Hi Aaron, for acquisition, a kind of hybrid 422 codec: 10b Y' and 8b C'b/C'r would be perfect. Cheers, Rafael

August 21st, 2009, 07:46 AM	#42
Aaron Newsome Major Player Join Date: Jul 2009 Location: Rancho Santa Margarita Posts: 590	This is why my workflow includes getting the rough cut on the timeline in FCP and immediately using media manager to create new clips for me, in a new project. Either 10 bit Uncompressed or 10 Bit ProRes422 HQ. Once I have the new project, I'm confident that the quality will remain through compositing, color grading, etc. You need to make sure you check "Delete Unused Media" in media manager though, otherwise it will export the complete length of your original clips.

August 23rd, 2009, 02:58 PM	#43
Barlow Elton Major Player Join Date: Mar 2005 Location: UT Posts: 945	Yup, I think rough cutting in an XDCAM timeline, and then using the media manager to transcode to ProRes for final edits/corrections/outputs, etc, is a great workflow. Much easier than converting everything to ProRes from the beginning. I think uncompressed 10bit is overkill for most anything though, especially if it's a lengthy project.