|
|
Posted by Smarty on 11/04/07 15:04
Regarding Penryn / Skulltrail:
Here is one very informative reference. Note the comparisons for TMPGEnc,
Virtualdub / DiVX mid way down the page in particular:
http://techgage.com/article/idf_07_sf_skulltrail_qx9650_45nm_performance_preview/
Smarty
"Smarty" <nobody@nobody.com> wrote in message
news:23lXi.1667$b%1.104@trnddc01...
> I'm really glad to hear the good news Paul, and my prior gushing
> enthusiasm hopefully now seems justified and doesn't seem induced by
> drinking special Kool-Aid.
>
> On the subject of HDV editing, I want to mention that Premiere Pro
> stumbles on machines which can otherwise do very competent and fast HDV
> editing, and I would argue that a 3 GHz Pentium 4 with a gig of RAM makes
> an absolutely fine editing platform with some software. Vegas 8 is
> actually pretty decent, and programs like VideoReDoPlus, a true bargain in
> the same spirit as the HV-20, is quite awesome. Ulead VideoStudio Plus
> version 11 is another cheapo wonder which handles HDV extremely well.
>
> If you are truly committed to running Premiere on a fast processor, for my
> money I would definitely await until November 12th for the release of the
> new Intel Penryn CPU, whose SSE4 instruction set speeds up video encoding
> by at least a 2X or greater factor when the programs have been written to
> use the new instructions. The current Core 2 Duo folks with 6600s, 6700,
> and 6850 processors will be very unhappy when their Conroe, Kentfield, and
> Woodcrest machines (including the 8 core MacPro Xenons) suddenly seem to
> be running at glacial speeds, comparatively speaking. Check out the
> "Skulltrail" links for more info. I assume that Premiere and others will
> be updated to exploit the new SSE4 instruction set. The DiVX encoder is
> already running SSE4 and is just blazing fast, a very promising piece of
> news for those of us also interested in AVC/h.264/AVCHD and all of the
> associated codecs.
>
> Good luck with the new HV-20. And definitely check out the Canon hi def
> wide angle converter.
>
> Best,
>
> Smarty
> "PTravel" <ptravel@travelersvideo.com> wrote in message
> news:5p4bg2Fpgn50U1@mid.individual.net...
>>
>> "Smarty" <nobody@nobody.com> wrote in message
>> news:OZ3Xi.437$m44.219@trnddc06...
>>> Paul,
>>>
>>> I just read your Saturday afternoon update on the DVI forum, and am very
>>> pleased to learn that readjusting the TV set sharpening apparently
>>> solved the problem.
>>>
>>> Today's overcast weather in the L.A. area may prevent seeing the HV-20
>>> at its' best, but I am anxious to see / hear how the experimentation and
>>> comparisons went. I meant it most sincerely when I said my 3-CCD FX-1
>>> could not compare to the HV-20, to such an extent that I literally put
>>> the FX-1 on craigslist.com and sold it once I saw how well the HV-20
>>> performs. Same thing with the HC-5. Others have made similar comments
>>> and reports.
>>>
>>> What are your present thoughts?
>>>
>>> Smarty
>>
>> I'm now a convert -- the HV20 is really amazing. For anyone reading this
>> who doesn't follow dvinfo.net, the problem I had experienced was a
>> combination of the camera's default setting for sharpening being too
>> high, and sharpening set too high on my high-def television. I really
>> don't understand Canon's thinking in this regard, as the default "medium"
>> sharpening setting in the HV20 really degrades high-frequency detail
>> without offering any advantage. Kudos, however, to Canon putting
>> sharpening adjustments in the user's control (I just wish there was a way
>> to turn it off entirely).
>>
>> I shot a few more minutes of test video a little while ago (it's sunny
>> here in Santa Monica) and the video is simply stunning. My VX2000 will
>> probably go on eBay tomorrow if I have the time to take some pictures of
>> it. I can't imagine why anyone would consider a different consumer
>> camera, and if you're willing to sacrifice some manual control (and an
>> XLR input), this thing compares favorably, actually more than favorably,
>> with prosumer offerings costing four times as much.
>>
>> There are still some minor motion artifacts, but no worse than I get with
>> VX2000, and I've been happy with that for years. The lens could probably
>> stand a few more elements -- there is some minor fringing in the
>> telephoto position but, again, nothing I can't easily live. I've also
>> figured out a way to put a strap on it that lets me manage my usual
>> shooting style of keeping the camera at chest level with the strap
>> providing support for the front of the camera. The HV20's image
>> stabilization is very effective -- the stuff I shot this afternoon isn't
>> as steady as when I shoot with a tripod, but it's pretty good and better
>> than what I could do with my VX2000.
>>
>> I haven't yet broken the news to my wife that I've bought Yet Another
>> Camera, but I think the combination of the tiny form factor, which she'll
>> like, and the fact that I can almost certainly get more for the VX2000
>> and all the accessories I have for more than I paid for the HV20 should
>> mitigate the coming storm when she finds out. ;) Actually, I'm looking
>> forward to doing some travel video without carting around the weight of
>> the VX2000, its extra batteries, the WA lens, etc. I barely notice the
>> HV20 when I'm carrying it, and the batteries weigh less than a fourth of
>> the VX2000's.
>>
>> The only bad news is that my 3 GHz P4 with 1 gig of ram isn't up to
>> editing HDV -- previews in Premiere Pro CS3 are jerky and often freeze.
>> I'm going to have to upgrade this computer (and I'm not looking forward
>> to telling my wife that!). Happily, though, my laptop, a Core 2
>> Duo-equipped Sony Vaio with 2 gig for RAM can manage editing HDV fairly
>> well. I'll probably use that for editing until I can afford to get a new
>> computer. One of the interesting possiblities it offers is to start
>> editing in the field, something I've never tried before. I'm way behind
>> in my editing projects, and still have at least one standard-def shoot
>> that I need to finish.
>>
>> The combination of 24p and "cine" mode on the HV20 is really interesting.
>> As I mentioned on dvinfo, I really have no need for this feature, but the
>> camera's ability to produce a fairly convincing "film look" is
>> impressive. If I was a young film maker or film student, I'd definitely
>> get one of these. Maybe I can talk Mrs. PTravel into considering a new
>> career. ;)
>>
>> I also went over to B&H's website this afternoon to order some
>> accessories. Of course, they don't take orders for another hour or so
>> (sundown in NYC), but they've got the HV20 for only $798! I assume that
>> means there's an HV30 on the horizon, but I'm off to Cambodia in a few
>> weeks and don't want to wait. Though I can think of some improvements to
>> the HV20, I don't think there's a compelling reason for anyone to wait.
>>
>>>
>>>
>>> "PTravel" <ptravel@travelersvideo.com> wrote in message
>>> news:5p12idFof3pjU1@mid.individual.net...
>>>>
>>>> "Spex" <No.spam@ta.com> wrote in message
>>>> news:13imfvhg31fbe63@corp.supernews.com...
>>>>> There doesn't look much wrong with this screen grab. There is a
>>>>> considerable amount of sharpening that will catch the eye as the video
>>>>> moves. With so much native resolution there is little value keeping
>>>>> the sharpening that high.
>>>>
>>>> Yep -- that's exactly what I'm seeing. I'm pleased with the resolution
>>>> and color fidelity of the image, but "shimmering" on movement is
>>>> driving me crazy.
>>>>
>>>>>
>>>>> PTravel, why not post a RAW m2t clip to Rapidshare??? It won't cost
>>>>> you anything.
>>>>
>>>> I'll try that this weekend.
>>>>
>>>>>
>>>>> How hot are the whites in the image? Check their level.
>>>>
>>>> I'll look this weekend.
>>>>
>>>>>
>>>>>
>>>>> Smarty wrote:
>>>>>> For whatever it is worth, the frame grab looks to my (60+ year old)
>>>>>> eyes as being pretty typical of what I would expect for the HV20 with
>>>>>> additional JPEG processing applied. If I look at the very finest
>>>>>> detail in the picture, such things as the parking meter next to the
>>>>>> car, small signage detail on the lamp-post as well as the very tip of
>>>>>> the lamppost, variegations in the fencepost masonry, and other really
>>>>>> small (just a few pixel wide / tall) elements, my impression is that
>>>>>> the capture has both preserved the edge definition without fringing
>>>>>> (unlike the cheaper Sonys which make artificial edge sharpness by
>>>>>> deliberate overshoot) and that there are no clear examples of a
>>>>>> defect. The areas you enlarged do look a lot different from the very
>>>>>> same areas which I enlarge using Photoshop, and thus your surrounding
>>>>>> magnified crops do look distorted but only when I look at your
>>>>>> magnified crops, and not when I look at the same high magnification
>>>>>> of these areas using the central 1440 by 1080 image. I am not sure
>>>>>> what to make of all of this.
>>>>>>
>>>>>> None of these comments / observations apply, of course, to what you
>>>>>> are seeing on your TV set, or what the moving, dynamic video looks
>>>>>> like versus this single static frame. All I am saying is that the
>>>>>> sensor and encoder appear to be doing their job for this image
>>>>>> without losing the fine detail, and that the detail which is
>>>>>> preserved does not (to my eyes) appear to have distorted or
>>>>>> exaggerated edges, color issues, or other evidence of a defective
>>>>>> sensor, encoder, or optics.
>>>>>>
>>>>>> You alone can judge how faithfully this HV20 is capturing the true
>>>>>> scene, and in this regard, another still camera with high resolution
>>>>>> might allow you to make some further comparisons. I will often use my
>>>>>> 8 MPixel Nikon to take comparison shots for seeing where the video
>>>>>> camera is weak. Obviously the color gamut and resolution is worse in
>>>>>> the video sample, but the comparison to a reference can help discern
>>>>>> the video shortfalls easier.
>>>>>>
>>>>>> It will be interesting to see if any of my observations agree with
>>>>>> anybody else's. I am not an expert at all in these matters, and have
>>>>>> cataracts to further confuse the issue, so I am merely offering my 2
>>>>>> cents worth in a sincere effort to be helpful.
>>>>>>
>>>>>> I'm glad to keep working this to get more insight into what you are
>>>>>> encountering.
>>>>>>
>>>>>> Smarty
>>>>>>
>>>>>>
>>>>>> <ptravel@travelersvideo.com> wrote in message
>>>>>> news:1193986489.385033.199060@q3g2000prf.googlegroups.com...
>>>>>>> On Nov 1, 7:51 pm, "Smarty" <nob...@nobody.com> wrote:
>>>>>>>> Paul,
>>>>>>>>
>>>>>>>> I have not seen the problem you describe, but it may be the fact
>>>>>>>> that I
>>>>>>>> virtually never view the camera through its' HDMI port. I normally
>>>>>>>> capture
>>>>>>>> the Firewire data / .m2t, edit and author an HD DVD, and watch the
>>>>>>>> resulting
>>>>>>>> (non-transcoded) output. When I have made direct comparisons from
>>>>>>>> tape
>>>>>>>> output from the camera's HDMI port versus playback of the HD DVD
>>>>>>>> via HDMI,
>>>>>>>> both had none of the "shimmer" you describe at the post on the DVI
>>>>>>>> forum you
>>>>>>>> linked to.
>>>>>>>>
>>>>>>>> I think the distinction you make regarding dramatic differences in
>>>>>>>> HDMI
>>>>>>>> versus component is very telling, however. Both should contain high
>>>>>>>> frequency components up to about the same cut-off frequency /
>>>>>>>> half-power
>>>>>>>> point. Each delivers essentially the same bandwidth and resolution.
>>>>>>>> And on
>>>>>>>> the 2 HDTV monitors I have here (both of which are 1080p) as well
>>>>>>>> as the
>>>>>>>> computers / monitors I have for editing, the progressive display
>>>>>>>> shows no
>>>>>>>> such effects. Perhaps the camera's 1080 interlaced signal is
>>>>>>>> managed
>>>>>>>> differently on your monitor when seen through the HDMI port versus
>>>>>>>> the
>>>>>>>> component input? I personally doubt the default setting of the
>>>>>>>> camera's
>>>>>>>> sharpening would (if the camera is working properly) create this
>>>>>>>> effect,
>>>>>>>> although reducing the high frequency energy with lower sharpening
>>>>>>>> may
>>>>>>>> "solve" the problem at the expense of the camera's excellent
>>>>>>>> resolution. If
>>>>>>>> it were me, I would want to see another HDTV / monitor with HDMI to
>>>>>>>> judge
>>>>>>>> how much of this, if any, is truly a camera issue. If it persisted
>>>>>>>> in the
>>>>>>>> second monitor, I would get a replacement camera.
>>>>>>>>
>>>>>>>> The mpeg encoder is imperfect, and motion does stress the encoder
>>>>>>>> so as to
>>>>>>>> make fast pans take on a more under-sampled and degraded
>>>>>>>> appearance, but
>>>>>>>> very very seldom have I seen any real macroblock effects or other
>>>>>>>> artifacting. My movies of Niagara Falls, with lots of vertical
>>>>>>>> water motion,
>>>>>>>> horizontal panning, and very agitated and tiny water details is
>>>>>>>> about the
>>>>>>>> most stressing case I have tried, and even then the HV-20 was a
>>>>>>>> vast
>>>>>>>> improvement over the older FX-1 as well as the recent Sony HC-3 my
>>>>>>>> son was
>>>>>>>> using.
>>>>>>>>
>>>>>>>> Although I am by no means an advocate of high priced cables, I
>>>>>>>> might also
>>>>>>>> take a look at the HDMI cable being used between the HV-20 and the
>>>>>>>> HDTV. The
>>>>>>>> cables I use here are very inexpensive, work very well, and
>>>>>>>> introduce no
>>>>>>>> particular problems, but it is remotely possible that high
>>>>>>>> frequency ringing
>>>>>>>> or other transient / overshoot problems could make the HDMI port
>>>>>>>> look bad.
>>>>>>>> This is a bit far fetched but worth a quick substitution if you
>>>>>>>> have another
>>>>>>>> HDMI cable to substitute.
>>>>>>>>
>>>>>>>> Please continue to update as I really would hope this problem is
>>>>>>>> not a
>>>>>>>> deal-breaker.
>>>>>>>>
>>>>>>>> Smarty
>>>>>>> Smarty, thanks for the response. I've got a bunch of HDMI cables
>>>>>>> here
>>>>>>> but, as it happens, the one I was using is the best I have -- though
>>>>>>> money isn't necessarily a measure of quality, this one cost me $70.
>>>>>>> I'll try a couple of others this weekend.
>>>>>>>
>>>>>>> I took a look at a frame grab. To my eye, there are sharpening
>>>>>>> artifacts or, at least, some kind of high-frequency distortion. I
>>>>>>> posted it here:
>>>>>>>
>>>>>>> www.travelersvideo.com/hv20.jpg
>>>>>>>
>>>>>>> In the interest of bandwidth, I had to compress it rather heavily,
>>>>>>> but
>>>>>>> I think the artifacts show through. I REALLY want to like this
>>>>>>> camera, so I'm going to do some more extensive tests this weekend,
>>>>>>> including lowering sharpening and trying it in 24p mode (though my
>>>>>>> preference is to shoot 1080i/60).
>>>>>>>
>>>>>>>> "PTravel" <ptra...@travelersvideo.com> wrote in message
>>>>>>>>
>>>>>>>> news:5oue5kFoaqqtU1@mid.individual.net...
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> "Smarty" <nob...@nobody.com> wrote in message
>>>>>>>>> news:J9aWi.29875$eD3.26430@trnddc03...
>>>>>>>>>> Looks like nappy is beginning to get tempted...... I am anxious
>>>>>>>>>> to see
>>>>>>>>>> how you and the other true professionals here find this camera,
>>>>>>>>>> since I
>>>>>>>>>> judge image quality and other related performance mostly as a
>>>>>>>>>> non-professional user.
>>>>>>>>> As everyone here knows, I am far from a professional, either in
>>>>>>>>> skill,
>>>>>>>>> experience or knowledge. However, I've hit a significant problem
>>>>>>>>> with the
>>>>>>>>> HV20 that may result in my returning it.
>>>>>>>>> Short version: there are significant motion artifacts in
>>>>>>>>> high-frequency
>>>>>>>>> detail, not unlike what you see with a Bayer-filtered single-CCD
>>>>>>>>> SD
>>>>>>>>> camcorder. The problem is dramatic on the camera's HDMI output,
>>>>>>>>> far less
>>>>>>>>> visible on component out. I'm still trying to figure out whether
>>>>>>>>> it's
>>>>>>>>> caused by over-sharpening in the camera, lousy HDMI circuitry, or
>>>>>>>>> something odd with my television.
>>>>>>>>> I've discussed it at length here:
>>>>>>>>> http://www.dvinfo.net/conf/showthread.php?p=768436
>>>>>>>>> Since these videos are for my own personal use only, I may keep
>>>>>>>>> the camera
>>>>>>>>> as long as it looks good on the component output. However, I'm
>>>>>>>>> going to
>>>>>>>>> do so more tests this weekend. If the output continues to display
>>>>>>>>> these
>>>>>>>>> signficant high-frequency motion artifacts, I'm returning it and
>>>>>>>>> waiting
>>>>>>>>> until next year to buy either an AH1 or FX7.
>>>>>>>>> I'm really disappointed -- I had high hopes for this machine.
>>>>>>>
>>>>>>
>>>
>>>
>>
>
>
Navigation:
[Reply to this message]
|