I don't know about "pq loss" (a term I am not familiar with), but I just raised these issues about tapeless formats in another thread:

Non-tape format issues

It's not the bitrate. It's the decode. The reason folks are having a hard time believing that you can be decoding AVCHD full framerate (in other words, seeing your preview window saying "best/full" and not dropping a frame, is because no one else can do it. On the fastest quad core system, I can't do it.
HDV is able to take advantage of optimization on the CPU that does not currently exist for AVCHD, although that is changing in a couple of weeks.
"Full time" means that you're seeing AVCHD coming off the timeline at a full 29.97, no dropping frames during editing, it's just a solid playback. In other words, so solid, you'd have full confidence sending it to air from the Vegas timeline as AVCHD, with preview set to best/full.
It's all about the codec, and how fast the system can decompress the encoded file, not the bitrate at which the file was encoded.
FWIW, I've yet to see any computer playback 18Mbps progressive files at full framerate, best quality, including quads, without rendering to RAM first.
If you're really able to do this, I think everyone that owns an AVCHD camcorder would like to know more.

BTW, as John Cline mentioned, there is zero difference recording AVCHD to DVD, MSPD, HDD, SDHC, or any other storage format, any more than there is any difference recording DV to an HDD or CF system, or recording HDV to an HDD, CF, or tape system. It's all just bits, at different bitrates.

Spot wrote:



What does the ability to preview the video have to do with the ability to successfully edit and render the video in the end? According to the manual, the preview has no effect on the final video.




8.0c?




According to the manual, the preview has no effect on the final video.



I have stated that what I have done up to now has not been an issue. I am eager to get home and attempt a 1920x1080 project.



If I have no issues creating a 1920x1080 project, I won't have an explanation as to why. The first assertions were that I would not even be able to edit an AVCHD project, now everyone is raising the bar in order to prove that I can't do it.

Even if you have a quad core you will probably have trouble. Drag your files direcet from the drive and you have MTS files.

You might as well plan on only opening a few at a time. I've had multiple crashes by trying to bring footage into a new veg.

Take your MTS files and convert them to something else like m2t that Vegas can actually work with.

Be prepared for a headache.
Laura

Laura K Wrote:
Even if you have a quad core you will probably have trouble. Drag your files direcet from the drive and you have MTS files.

Some people have written that Vegas does not play nice with quad cores and that could be your issue. I could check the exact number when I get home, but I was able to drag at least 90 small MTS files into my timeline, plus add a music score to this video without any crashes:

http://www.vimeo.com/843222


EDIT: 60 small MTS files

"Then why is everyone so friggin amazed that I can edit AVCHD files with my PC when AVCHD maxes out at about 17Mbs right now?"

Because AVCHD requires much more CPU horsepower to decode than MPEG2. Also, one would think that the more compressed the file is, the easier it will be to decode it. The truth is, the more compressed a video file is, the more horsepower it takes to decode.

I am sure dspenc1 can edit his AVCHD albeit slow. The preview has to be horrably slow as well.
I am running a Q6600 with 4Gig RAM, and only working with SD stuff. If I preview with best/full I am dropping frames left and right.
Can't imagine what my poor quad core would do with AVCHD....

Ted

There seems to be confusion on several points. Perhaps I am confused as well (what else is new?), but here's my attempt to address some of the muddy stuff:

1. What does the ability to preview the video have to do with the ability to successfully edit and render the video in the end? According to the manual, the preview has no effect on the final video. I think the point Spot was making is that you don't really know if you have the video you really want, unless you can preview it at full quality ("Best" preview) and at full frame rate. You are correct that even if you can only preview at 2 fps, draft mode, that the video will still render correctly, but who wants to have to render to a final format and then watch on your target monitor, just to be able to tell if some fX tweak worked? So, full-quality, full-framerate preview is essential for most people's workflow, and I think the ability to be able to do that is what people are really questioning.

2. Also, one would think that the more compressed the file is, the easier it will be to decode it. The truth is, the more compressed a video file is, the more horsepower it takes to decode. I think this may be the first time I have disagreed with anything John has written. Any form of compression -- whether lossy or lossless, and whether generic (like Zip) or specific (like JPEG, or MJPEG, or MPEG-1, MPEG-2, HDV, DV, AVCHD, etc.) is asymetric meaning that it usually takes more CPU horsepower to encode the video than to decode. This is especially true for compression that uses interframe compression, where the encoder must look at a dozen or more frames in order to create one frame of video. To do this, the encoder must "try" things and then make decisions. By contrast, decoding is a very linear process, where the decoder just follows rules, "turns the crank," and the video appears. The amount of CPU power required, however, doesn't have anything to do with the amount of compression, but instead depends on the algorithm used and the number of frames that must be referenced.

Now, here's the main part of the post. The amount of compression is determined by the actual math used (like DCT) and also on the compression factor. However, whether you compress by 10% or 99% (by throwing away more stuff), the math is identical, and therefore it takes the same CPU effort to decompress, and you end up with the same number of bits per frame (a 720x480 video frame is the same whether it was created from uncompressed or from 99% compressed video). In fact, the more compressed, the fewer the number of bits that must be read from the hard drive, and therefore the more compressed the video, the faster it will decompress, although the speed improvement from having a smaller file is generally not a huge factor. However, sometimes it can make a big difference, for instance when trying to open a JPEG file with no virtually no compression over a slow network connection. The time required to read the file can be quite large. By contrast, take the same file and compress it down to next-to-nothing, and it zips across the slow network connection, and then the CPU decompresses it in basically the same time as the larger, less compressed version.

"Wow, so much anger ..."

Where the heck do you see anger???

In fact my wife was sitting on the couch wondering why I was smiling when I wrote this.

Why are strong words always construed as anger? It's merely the way I talk.

It's really funny how people jump to conclusions. I remember one post was all done in capital letters and someone asked why all the yelling. The poster asked why it was assumed that he was yelling and they replied "because of the capital letters". The poster replied "but I'm not yelling... I merely don't see that well."

OK. I redid a project using 1920x1080 settings. I also rendered at 1920x1080 using the Best rendering setting.

I changed out the soundtrack and edited the title. In previewing the project on Best/Auto, there were some dropped frames. It took a total of about 25 minutes to render the video, about a 18 to 1 ratio on the length of the video. I played back the file in Apple Quicktime and roughly half the frames were dropping. My monitor does not support 1920x1080 resolution, so much of the video was off screen.

I have no need to create 1920x1080 video using an AVCHD file as the source, but it can be done on my wimpy system. Playback is another issue.

The video can be seen at http://vimeo.com/880992. If you are a Vimeo member, you will be able to download the file.

I hope this puts the issue to rest.

I see you, but who does the hand belong to?...

All it puts to rest is what we're all talking about; you're not able to preview AVCHD files on the Vegas timeline at full framerate.
Whether you can or can't render a file is not a question; of course you can.
For most folks that edit video, being able to see smooth/full framerate in the Preview window is a very big deal. If you can't, it's not considered a viable edit.
Sure...if you don't mind jerky playback during the editing stage, then its working great for you. Most folks here don't recall when DV was exactly the same as what AVCHD is today; it required an external, hardware decompressor. Now we edit DV with nothing but a harddrive and a cheap laptop. AVCHD is the 'new' DV, and it's very problematic to edit at full framerate, and at any resolution.
But rendering...that's no big deal. Rendering to a 1920 x 1080 AVCHD file is no big deal. Playing back that rendered file in Quicktime or other player is no big deal, depending on the bitrate at which itwas encoded. But editing with the same smoothness and framerate as it's seen in Quicktime...that is a big deal. And so far, no "cheap" machine, and most quad-core machines cannot play back AVCHD on an NLE timeline. You've more or less confirmed this. And in the process, let folks know you don't have some magical elixer that everyone else in the world is missing. :-)

"HDV is able to take advantage of optimization on the CPU that does not currently exist for AVCHD, although that is changing in a couple of weeks."

"AVCHD is the new DV."

Statements that send shivers down my spine. With apologies due to having very little info on AVCHD, does this mean that if I'm taping in HDV my quality has already been bypassed and is inferior to AVCHD? I thought I was on the cutting edge taping with my V1U... well, at least compared to the SD that I see local videographers at my [low] level using.

"HDV is able to take advantage of optimization on the CPU that does not currently exist for AVCHD, although that is changing in a couple of weeks."


So could somebody enlighten me and tell me what will be happening in a couple of weeks ????

Your HDV is not obsolete, not for a long while. However, camera companies are moving away from tape based solutions towards AVCHD other disk formats.

You will still be able to buy tapes for a long, long, time.

With apologies due to having very little info on AVCHD, does this mean that if I'm taping in HDV my quality has already been bypassed and is inferior to AVCHD? Do a search on DSE or Spot and include AVCHD in the search. He has posted quite a number of times about the differences between HDV and AVCHD. I believe that his experience has shown that AVCHD does not hold up anywhere near as well as HDV. Here is a sample of one of his posts:

HD Camcorder Purchase Question

AVCHD is newer, but that does not necessarily make it better. It is aimed at solving a different set of engineering and marketing propositions, by using faster CPUs and a more aggressive compression scheme.

Some of us may not realize (I didn't till I saw a DSE seminar a few years ago) that DV is a lossy compressed format - and that is what a lot of us start with as source material. To reiterate what he was saying, in the early days of DV (and I remember my early days with it, in 2000) it was a "new format for camcorders" and it was completely digital, etc etc. AND, it was difficult for computer systems and software to keep up.
I think what he is saying is that these days, we're experiencing similar issues with AVCHD. It's a new (HD) format for camcorders, new solid state media, new compression, and it's difficult for computer systems and software to keep up.

All other things being equal, today, I think I'd much rather have HDV on tape.

HDV is a much better way to go, but even in the early days of DV, it was far easier to deal with than HDV or AVCHD is today. Heck, I started on a 450MHz Pentium, and it captured without dropping frames, and was able to playback (using that old buggy Pinnacle Studio) at full frame rate. Wow! By contrast, people here are talking about computers with multiple processors, with multiple cores, multiple threads, and processor clock rates almost an order of magnitude faster, and memory that is 20x what I had in that old computer (256 MB, if I remember correctly).

My point is that given that we've had these long-GOP formats for several years, I don't think they are likely to ever been as easy to deal with in native form as DV. The solution, for those who want the "good old fashioned" performance of DV is to spend the money and take the time to generate intermediates and deal with those instead. Near as I can tell, despite some problems that Cineform development and Sony development have had in coordinating their upgrades, once you get this working, it is pretty sweat. All it requires is the extra money to purchase the Cineform, and some extra drive space for the 5x increase required for the intermediates.

Just to clarify what I meant;
HDV is superior (currently) to any AVCHD camcorder available. AVCHD is a MONSTER on the CPU, and in spite of the BS that some manufacturers spout about AVCHD being "twice as efficient as long gop MPEG 2" (and it is, at very low bitrates, as you'll see I posted a couple years ago) it is not the great thing that some would have you believe.
That having been said, AVCHD is the "new DV" because it's consumer accessible, like DV has been. HDV is a superior format, but it's not "consumer accessible" like the bottom end DV was/is.
It'll be a while til we see the efficiencies of AVCHD match the reasonably fast efficiency of HDV.
AVCHD offers new challenges. It's a decent format, I'm using it regularly on the weekends for skydiving video, but it brings a quadcore to its knees when it comes time to color correct and render. It's horrid to downconvert to SD without some serious horsepower.
There are alternative workflows, however, and more and more AVCHD optimized tools will be seen in the very near future.
Treated properly, AVCHD is a terrific acquisition format. But right now, it's a bear post-acquisition.

That's good to hear re pre- vs post-acquisition w/AVCHD ... thanks for the insights.

I just ordered an SR11; so I'll see how it works on my dualcore e6660 4 gigs ram winxp. I really like the idea of a tapeless workflow, that's integrated re no external drives needed to capture footage, as that will simplify and speed up workflow, at least on the acquisition/conversion to usable by pc format, end of things.

Looking forward to seeing what AVCHD optimized tools are available soon, to help with the post-acquisition editing/rendering end of it. I'm planning on still using SD NTSC size projects in Vegas (or 480x360 for web-to.flv format), hopefully when I drop the larger AVCHD footage into the timeline, I can somehow still get fullscreen previews, if I'm focusing on just part of the footage frame, and/or resizing it within Vegas, or some other workaround, vs dropping preview frames.

-k

Spot wrote:
HDV is able to take advantage of optimization on the CPU that does not currently exist for AVCHD, although that is changing in a couple of weeks.

Spot, what do you mean things will change in a few weeks? If you can't elaborate, are you at least suggesting that anyone considering a purchase should wait a few weeks?

VASST and NewBlueFX released AVCUpShift at NAB. It rapidly converts AVCHD to a more comfortable, faster codec.
Price is 49.00 for the app. Allows conversions as high as 60Mbps. It's VERY fast to edit.