February 21st, 2014, 05:19 PM | #331 |
Inner Circle
Join Date: Mar 2003
Location: Ottawa, Ontario, Canada
Posts: 4,222
|
Re: Sony FDR-AX100
60P and 60i have exactly the same temporal motion they both record 60 ( 59.94) exposures a second. 60i records fields, half the vertical resolution, and 60p records the whole frame. Confusion arises over the timecode as 60i is 29.97 fps , since the timecode increments every 2 fields.
As mentioned in a previous post and modern interpolating display like my 240Hz Sony can make 24P, 30P, 60i and 60P all look much the same by filling in the missing information to display progressive frames at the refresh rate of the display. In my case an early true 240hz Sony For my use 30P will not do it. Ron Evans |
February 21st, 2014, 05:53 PM | #332 |
Trustee
Join Date: Aug 2010
Location: Florida
Posts: 1,197
|
Re: Sony FDR-AX100
I'm perfectly fine with 29.97p. It's my favorite frame rate with my EX1r and FS100 anyway.
I think 29.97 handles motion perfectly fine. (significantly better than 24p) 29.97p video also creates a perfect progressive "look" when it's rendered out to 60i for Blu-ray too. Because both top and bottom fields are sliced from the exact same frame, you get no "comb" effect in 60i. It's a basically 29.97 PSF. So no 60p on the AX100? No problem for me. I'm more curious to see how assigning only 15Mbp/s for each 1080 quadrant will look. 4 x 1080 = 4K or almost UHD anyway. XAVC-S at 60Mbp/s? Hmmm.....big question! |
February 21st, 2014, 06:33 PM | #333 |
Inner Circle
Join Date: Feb 2007
Location: Apple Valley CA
Posts: 4,874
|
Re: Sony FDR-AX100
To my eye, 60I and 60p look very different in practice, and 24p "stutters" for want of a better word - looks jerky to me, vs. "smoother" 60p. Philosophically, I want "more data" at the time of capture if possible, with the theory that it gives more to work with later... so the idea of 30 vs. 60 is not appetizing.
BUT for some time I've rendered out to 24p from 60p with good results... there seems to be some secret sauce involved in the processes that is not necessarily expressed in the specifications... as we've discussed a bit here, there are "specs", and there are results. We all saw the early RX10 "test shots" from STILL CAMERA reviewers that were a little scary at best... now that people who understand that swinging the camera wildly is NOT "technique" are using it, the RX10 is turning out to be a pretty good VIDEO camera, as designed. I guess this is why I'm holding out some hope that 30p 4K may actually turn out to be "useful" or usable... despite the initial reaction that it won't cut it, based on "prior experience". We will see soon enough, and it'll likely still be pretty impressive for 1080/60p no matter what (OK, so save $500 and buy the CX900?). In that respect, the question is whether if's a step "up" from the CX/PJ7xx series that so many of us know, and use for what they do so well. |
February 21st, 2014, 06:57 PM | #334 |
Major Player
|
Re: Sony FDR-AX100
I'm fine with 29.97p too. I shoot at 59.97p with my RX10 and when I render off at 29.97 I can't see any difference (vs 60p). Maybe it's my subject matter (or my eyes).
The conversation about different needs is interesting. For what it's worth here is my list. 1 A decent EVF 2. Noise-free sharp colourful 29.97 or 59.97p 3. Long (optical) reach 4. Light weight, compact, highly portable (not necessarily pocketable). 5. High quality image stabilization 6. RX10 quality stills 7. Built in ND filters 8. Perfect auto-focus and auto WB. I don't need raw, hdmi out, 24p, shallow DOF, low-light capability, XLR audio inputs, touch screen. Along with my RX10, I think if I get an AX100 and an HX400v I might be set, for now! Until something better comes along, and it will. |
February 21st, 2014, 08:19 PM | #335 |
Regular Crew
Join Date: Mar 2008
Location: Tampa Bay Area, FL (USA)
Posts: 142
|
Re: Sony FDR-AX100
Thanks Ron & Dave for reminding me to look at the numbers. I completely didnt even think about the fact of resolution being 4x bigger than HD. After a while all the numbers start to hurt my short attention span prone brain. hahaha on a more serious note, like many others have mentioned, I'll have to get my hands on a camera and play around to see what will fit my needs and work style.
I'm also following discussions on the AX1 and am keen to find out just how "compatable" the AX1 and AX100 will be with each other as I'm sure others are as well. |
February 22nd, 2014, 03:31 AM | #336 |
Regular Crew
Join Date: Mar 2007
Location: Kaapstad South Africa
Posts: 64
|
Re: Sony FDR-AX100
Not sure, if this video is shot with a Sony 4K cam, but i don't like the image at 1:20 (the higher speed images).
Or it maybe a crappy conversion. |
February 22nd, 2014, 08:12 AM | #337 |
Regular Crew
Join Date: Sep 2009
Location: Millstone,NJ
Posts: 194
|
Re: Sony FDR-AX100
Another video showing it must have been a long FL lens on it for many of the shots.
|
February 22nd, 2014, 08:51 AM | #338 |
Inner Circle
Join Date: Mar 2003
Location: Ottawa, Ontario, Canada
Posts: 4,222
|
Re: Sony FDR-AX100
I have had my FDR-AX1 since just before Christmas and have mainly shot 60P but a few shoots at 30P 100Mbps for a YouTube video that had no real movement anyway. Most Sony's match each other very well so I expect the FDR-AX1 and FDR-AX100 both at 30P same data rate to match well. I think low light will depend on zoom position for both cameras as the FDR-AX1 full wide is F1.6 but with both zoomed tele they are both F3.4 ( I think ) so in this case the FDR-AX100 will have the advantage.
Set wide for full stage view next to my CX700 the FDR-AX1 performed very similar both going to 21db of gain at times. The difference is when the FDR-AX1 is downconverted to 1920x1080 it seems to have more depth to the image likely because the colour resolution is now closer to 4:2:2 than 4:2:0 of the AVCHD from the CX700. For highlights this seems to add some detail lost in the CX700 image. Big advantage when shooting dance with flashing lights of all colours. For a dance show I left it full wide then cropped in to the stage area as needed and this worked great. Would love it to have better low light but it works for now. Ron Evans |
February 22nd, 2014, 09:01 AM | #339 |
Inner Circle
Join Date: Mar 2003
Location: Ottawa, Ontario, Canada
Posts: 4,222
|
Re: Sony FDR-AX100
Dave, watching 60i and 60P is totally dependent on the playback chain. 60i deinterlacing is different on different TV's so will look different between manufacturers etc. 60P is native for most flat panel displays in North America so will be more consistent across manufacturers. If from a DVD or Bluray player one has another option involved. Hence 60P is the best choice but of course no spec for discs just as there is no spec for 30P either !! So we are at the point where discs do not playback the two likely choices for people to shoot in !!!! Both have to be converted to 60i or 1280x720P60. I think we will see a new disc spec for 4K that includes both 60P and 30P !!!
Ron Evans |
February 22nd, 2014, 12:27 PM | #340 | |
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: Sony FDR-AX100
Quote:
Firstly, is the 60Mbs an average or peak value? I believe the numbers quoted for AVC-HD are peak - the average bitrate is much less than 24Mbs. And a lot depends on the individual coder - all AVC-HD coders don't give anything like the same performance, even at the same bitrate. And here we're talking about X-AVC which *should* give better performance than AVC-HD. But most of all, it's incorrect to directly scale compressed bitrates up as resolution increases, assuming equivalent quality. As a rule of thumb, the higher the resolution, the higher the compression ratio for equivalent quality (assuming all else - codec etc equal) So - the question you put is "is 60Mbs good enough"? That can only be really answered when we see real results - but even then, what is "good enough"? That may seem a silly question, but as with HD codecs, it depends on factors such as how much post processing the signal is likely to have. Which is why AVC-HD for HD may be fine for many uses - but is not broadcast approved. The original pictures may be fine - but can't be guaranteed to stand up to the compression/recompression of the broadcast chain. My gut feeling is that it probably will be OK - bearing in mind we are talking about a consumer camera. (Albeit a quite expensive one, aimed at the top end of "consumer".) Look at it this way. I believe top end cameras (such as the F55) code 4K with XAVC at a rate of 10Mb/frame - so 24fps is 240Mbs, as it's an I-frame only codec. Here we're talking about a codec a quarter the bitrate - but with the efficiencies that long-GOP brings. The latter may not fully compensate for the extra 4x compression - but we are talking about a consumer codec versus a true pro one. It may be tempting to think "but 100Mbs could only be better, surely?" Maybe true in terms of quality, but think of the designers job of reconciling design compromises. One of those is GUARANTEEING reliable performance as much as possible with consumer media - which argues in favour of keeping bitrates down. I'm sure Sony could have implemented 100Mbs - but it would have caused problems with guaranteed media etc to a far greater extent than 60Mbs recording. We must never forget it is a consumer camera. |
|
February 22nd, 2014, 01:36 PM | #341 | |
Regular Crew
Join Date: Feb 2009
Location: Los Angeles, CA, USA
Posts: 148
|
Re: Sony FDR-AX100
Quote:
I think the above will also happen with the AX100. You can bet your bottom dollar someone will use it filming landscapes outdoors with undulating water in the background, changing light, trees swaying in the wind, smoke in the scene with fast camera pans and deem the camera garbage because it's 60mbps couldn't cope with that stressful codec situation. Others will latch onto that and repeat it over and over again on various forums even though they have never even used the camera, but it was garbage for that one user therefore just like the VG900 it must be garbage for everything. Meanwhile my needs to the AX100 are 99% indoor filming with controlled light that doesn't change like outdoor lighting can, no visible smoke or water that tends to stress codecs, mostly static scenes that don't have all sorts of organic matter swaying in the wind, with normal speed pans. In my case 60mbps may do just fine. So for one user 60mbps is garbage, for me 60mbps may work ok. In the end it really does depend on how people intend to use the camera, for some the 60mbps code may not work but for others it may be just fine. It's not really something looking at pure numbers will tell you, it will depend on how the camera will be used. |
|
February 22nd, 2014, 05:05 PM | #342 |
Inner Circle
Join Date: Feb 2007
Location: Apple Valley CA
Posts: 4,874
|
Re: Sony FDR-AX100
@Peter -
As you point out, a camera that works for one person/job may not be "right" for someone else or a different job! It's important to remember that there isn't a "one camera fits all" - thus why we all might have a (ahem) "few" cameras... I have found the RX10 and RX100M2 have put many of my "favorite" cameras up for review for possible sale! I sort of expect the AX100 will complete the "sweep", and cover much of what I will "need" a camera/cameras to do... I'm sure there will still be a couple cameras that will be sticking around, but get used less... it's hard to compete with small, light, "fun" cameras that shoot really good stills/clips... As far as bitrates - if we rough calculate that 30p would require about "half" the bits of 60p, and current Sony "high end" bitrate for 60p is 28Mbps... at least in theory, I would think that 15Mbps per 1/4 screen in 4K should be able to "match" current 1080/60p in terms of individual frame quality... it'll come down to whether motion looks "right" or not - general opinion is that 28Mbps holds up "OK", but not "great" with higher motion... we'd all like "more" (thus the new higher bitrate 1080/60p in the CX900/AX100 may turn out to be a hidden gem), but there are limits... As the other David alludes to... In the end, digital is just 1's and 0's, artfully arranged, with varying degrees of success. Processing capacity and speed, as well as storage capacity and speed govern how much can be done at economical prices - those 1's and 0's have to be juggled and stored somehow, and over time we get more proficient/efficient ways to deal with more of them! At a $2k price point, you're still talking about HALF of other possible 4K "solutions" announced (presuming you'll have to buy into lenses, etc for a GH4...) or available (AX1)... with the pedigree the sensor and processor are already bringing to the table, this should be a breakout camera in many ways. High price always "feels expensive", but when you break down the features and capabilities, sometimes the price starts to look a lot more "reasonable" (OK, it's still a lot of money... but sometimes you actually do "get what you pay for"). |
February 22nd, 2014, 07:51 PM | #343 | ||
Inner Circle
Join Date: Jan 2006
Posts: 2,699
|
Re: Sony FDR-AX100
Quote:
It's one thing that a doubling of resolution does not mean a doubling of bitrate for equivalent quality, but doubling framerate most certainly does not mean doubling bitrate. It takes advantage of there being more redundancy in the higher framerate signal, and normally involves keeping the time interval between I-frames constant - typically every 1/2 second. Hence, for 30p there will be 15 frames in a GOP, for 60p there will be 30 frames. But in terms of size, the I frames are far larger than difference frames - and there is still only one per GOP, or two per second. Yes, about twice as many difference frames - but they are small in comparison, so overall the effect is far less than a doubling of bitrate. Quote:
And that's the problem with most "real world" tests. Unless they are done under controlled conditions, they are frequently meaningless - and often judgements get made on the quality of the photography, the lighting, the editing, and the subject matter! Not the quality of the camera - which is what is supposed to be under test! |
||
February 23rd, 2014, 12:05 AM | #344 | |
Trustee
Join Date: Aug 2010
Location: Florida
Posts: 1,197
|
Re: Sony FDR-AX100
Quote:
I agree with you on all of this. With a long GOP codec, it's very very tough to estimate and translate bitrates and quality between HD and 4K. It's not as simple as just dividing UHD video by a factor of 4 and comparing it to 1080. To make things even more complex, AVCHD runs on H.264 at a "4.0" level. This means that is does NOT contain all the H.264 tool sets that the h.264 spec has to offer. On the other hand, XAVC-S runs at "5.2" level. This means that XAVC-S has ALL the possible math and tools that h.264 allows. (5.2 is the most complex specification) So,...we can assume that bit rate for bit rate and pixel for pixel, XAVC-S will be better looking. I'm still skeptical but the only way to know for sure is shoot and test. lol I'm willing to give it a fair chance. I'm hoping it's fairly durable in fast motion, complex scenes and hopefully has fairly clean blacks. If the bitrate is too low, it will show up as heavy quantizing in the shadows first . Anybody shoot 4k at 60Mbp/s XAVC-S with their AX1? CT |
|
February 24th, 2014, 09:59 AM | #345 |
Inner Circle
Join Date: Mar 2003
Location: Ottawa, Ontario, Canada
Posts: 4,222
|
Re: Sony FDR-AX100
Cliff I will try and shoot something this week. I have not shot anything at 60Mbps 30P. It is -10C outside at the moment and just snow etc so not ideal for shooting. You will need some movement to test so indoor static is not of much use but could do a slow pan I suppose.
Ron Evans |
| ||||||
|
|