D3/D300: Myths and Facts -- Thom Hogan's Q&A

[arthuryeo]

Some nice analysis and presentation of facts from our friend Tom Hogan ..

 

http://www.bythom.com/D3QA.htm

[tonybeach]

It's news to me that the D300 will only shoot 2.5 fps using 14 bits color depth (technically, luminance depth). I clearly will use that feature sparingly, besides the fact that it's going to make the files larger (I'm guessing about 20 MB a piece; but IIRC, Fuji's 12 MP, 14 bit files are 25 MB). I'm pretty sure there will be uncompressed options for NEF, otherwise that would represent a small step backwards from the D200.

[ilkka_nissila]

I find Thom's comment on the RAW compression odd. If the menus have options for lossy compression and lossless compression separately one would naturally conclude that there is an uncompressed option as well.

[ShunCheung]

First of all, it is T<B>h</B>om Hogan; he doesn't like to be called Tom.

<P>

Nikon USA has confirmed that the frame rate would drop if you shoot 14-bit NEF on the D300, but they didn't specify the 2.5 frames/second part.

<P>

BTW, from what he writes, it is obvious that Thom Hogan is quite sore that Nikon USA didn't invite him to the product launch in Japan, the one Bjorn and Ellis attended.

[Ed_Ingold]

14-bit files are no larger than 12-bit files. Any digital integer between 8 bits and 16 bits is represented in a 16-bit word.

 

Regarding invitations, neither Bjorn nor Ellis ever intimated they were on the "inside" track. Thom is a competent writer, but tends to indulge in speculation. In this regard, I suspect Ken Rockwell would be stopped at the border, let alone at the the gate.

[ShunCheung]

The product launch in Japan was by invitation only and the subsidiaries at each country invites their local photographers. Given that there were some 350 people invited, I am sure Nikon USA had a fairly large quota after Japan itself. For example, Rob Galbraith was there, among many others:

http://www.robgalbraith.com/bins/content_page.asp?cid=7-8742-9087

 

Ellis is actually now a Canon user, despite that he continues to post in this forum. But I see Nikon USA is trying hard to get Ellis back. :-)

 

Given that Hogan has written a lot of books and eBooks on various Nikon products, it should be in Nikon's interest to keep Hogan informed. I think the problem is that Hogan is sometimes very critical of Nikon and certainly engages into a lot of speculations and leaks, while protecting his source.

[jeroen dommisse]

And what's Tom's fixation with the VR? Surely it's not that important, at least not in the 12-24 and 24-70 full frame range.

[bjørn rørslett]

I don't consider being critical to Nikon products was held against anyone. I for one was invited :) and in fact by invitation of Nikon Europe BV, a subsidiary of Nikon with which I have had my fair share of disputes.

[ellis_vener_photography]

No Thom wasn't on the trip --which surprised me -- and neither was KR --which didn't -- and he didn't see everything we did or talk withthe people we did. Off the record some forthcoming stuff was talked about but nothing specific and certainly nothing I'm going to repeat in an open forum. Bjorn may have heard the same stuff but I don't know who he spoke with as for the most part we saw each other only during meals or at the big press conference. Bjorn did let me know that I was doing a wonderful job of slaughtering the musical qualities of the open vo

 

Shun, there were roughly 20-25 people there in the USA/Canada contingent including Rob Galbraith, Michael McNamera (Pop Photo), Christopher Robinson (Digital Photo Pro), David Schloss (Photo District News /Aperture Users professional Network), and the truely impressive David Etchells of http://www.imaging-resource.com (now there's a guy who really knows his technical stuff). There were also editors from computer magazines and business analysts. I didn't have a chance to meet everyone.

 

Regarding a higher pixel count version of the D3. Maybe, but it wasn't discussed. I suspect there there are far more wildlife and hybrid wildlife /landscape photographers out there using DSLRs then there are pure landscape photographers using DSLRs. And the serious wildlife guys are going to absolutely love the D3.

[mark_chappell]

<I>And the serious wildlife guys are going to absolutely love the D3.</i><P>

 

As a 'serious widlife guy', I think the bigger news from Nikon is the VR superteles. And

for wildlife work that requires high frame rates, the D3 is MUCH better than previous

Nikon offerings. However, it's perhaps worth noting that for a

lot of wildlife (= telephoto) work, a sub-frame DSLR has advantages over full-frame.

That's

one of the reasons the 1D series has been a big success on the Canon side. The D3 is

certainly a

big

plus for sports shooters, and for those who like to shoot wildlife AND also are fond of

landscapes, but a high FPS DX format 'pro' machine (D2X on steroids) with improved noise

characteristics -- probably not feasible

with small photosites -- would be even better.

[gman]

That is a bummer that Thom Hogan was not invited. I have found his website and books to be very helpful and informative.

 

But hey, I was not invited either! :] I even offered to be a sherpa to some of those invited. I would have provided my own meals too. Oh well!

[ShunCheung]

As I mentioned before, as soon as I found out that both Bjorn and Ellis were in Tokyo, I quickly informed them both that the other one was also there, and I checked with Thom Hogan. He quickly replied that he wasn't invited (to my surprise also) and his displeasure was obvious.

 

I have been saying for a while that having a 4MP sports DSLR (D2Hs) in 2007 was an embarrassment on Nikon's part, and I am sure they don't like to read this comment. That is the one area Nikon should have fixed 2, 3 years back, ever since Canon introduced the 8MP, 8 frames/sec 1D Mark II back in early 2004. That is how far behind Nikon is until the D3 is available.

 

As a long-lens user, I have mixed feelings about the D3 having the FX format. Having more pixels cramped into a DX sensor lets us use shorter lenses or get more reach. However, having only 12MP spreaded on an FX sensor allows larger photosites and persumably better low-light results. That can be a major plus for indoor and night-time sports photography. Apparently, Nikon decided to optimize that way.

 

Canon's top 3 DSLRs are the 1Ds Mark III, 1D Mark III and 5D. So logically, Nikon should move to a high-pixel (1Ds) and another "prosumer" (hope you don't mind that term again) FX DSLR similar to the 5D. But I am sure Nikon will try sell as many D3 as they can for now before they introduce any "economy" FX DSLR.

 

Whild we are on the topic of Thom Hogan, he has another interesting article:

http://www.bythom.com/D3comments2.htm

 

And we have referenced this excellent report by Dave Etchells (mentioned by Ellis) a couple of times already, in case you missed them:

http://www.imaging-resource.com/NEWS/1187901361.html

[ilkka_nissila]

Since the fps and autofocus on the D3 and D300 are very similar, an action photographer would use the D300 for telephoto work where reach is critical, and the D3 for low light work or where you already have a long enough lens in the 35mm format.

 

And we can be pretty sure that a D3X with 20+MP in 2008 will cost around 7-8k so you can actually get both a D3 and D300 for less money!

[thom_hogan]

> I'm pretty sure there will be uncompressed options for NEF

 

The specifications list this now. But the UI seems overly complicated.

 

> it is obvious that Thom Hogan is quite sore that Nikon USA didn't invite him

 

I'm not sore. Sore would be an ego-based reaction, and I'm not interested in boosting my ego. I do think that NikonUSA's choices of invitees miss somewhat on the key influencers in the Nikon market, and this points to marketing still not being completely attuned to the market.

 

Moreover, a good marketing and PR approach is 100% inclusive. If you're not going to invite someone to an event, you still have some sort of outreach to everyone. That would mean, for instance, that coincident with the event you send out the PR to everyone who's an influencer and make them feel "part of the team." Nikon did not do that. As I've written before, despite asking many times, and despite a promise from the CURRENT agency doing Nikon's US marketing, I do not get press releases from Nikon. Am I sore about that? No. I think they're simply inept. Even an email to the press center saying the photos are now available would be a step up from their current functioning.

 

> Thom...tends to indulge in speculation

 

That's funny. Had Nikon invited me to the introduction I would have had to sign an NDA that would have precluded me talking (although vaguely to protect sources) about the products before they were announced. FWIW, I honor all NDAs I sign. And I'm not under NDA from Nikon. Thus, I'm free to use the material I regularly get to speculate on future products. You might note that I was talking about an Osaka Games launch of the D3 for months. That actually wasn't speculation ;~).

 

> it should be in Nikon's interest to keep Hogan informed

 

Here's the latest: my NPS Priority Purchase application for the D3 and D300 was rejected by NikonUSA as "premature." No real indication to me when it wouldn't be "premature." But I find it very strange that I can preorder a D3 and D300 from Ritz and other sources but not from NikonUSA itself.

 

> what's Tom's fixation with the VR

 

Thom's not fixated with VR. I specifically said that I think that NOT having VR on the 24-70mm was sending a mixed message to customers. It's the old Canon/Nikon thing again. What was one of the criticisms that would be used against Nikon for use in sports/wildlife? The long lenses didn't have VR and Canon's did. Canon has stabilization in mid-range zooms, Nikon doesn't and has just sent a signal that it WON'T. That's a little akin to shooting yourself in the foot (again). Even in the consumer mid-range zooms we've only got one with VR. If you're trying to read between the lines you come to one of two conclusions: (a) Nikon is slow in getting VR into their lineup; or (b) Nikon will have sensor-based VR some day. Neither of those are the right message to send right now, especially with Sony advertising "Our competitors are Shaking...".

 

> the serious wildlife guys are going to absolutely love the D3

 

I'm not convinced of that. Pixel density is one of the key attributes of a wildlife camera because you are almost always limited in approach to animals in the wild. The D300 is a far better choice in that regard. Moreover, the AF covers a broader range of the frame on the D300. The only question is whether the D300 has lower noise and higher dynamic range than the D2xs, in which case it absolutely replaces the D2xs as the wildlife camera of choice.

 

-thom

[ShunCheung]

Illka, I too would imagine that there will be some sorts of "D3X" with more pixels to compete against the 1Ds Mark III as a studio camera, but I am curious about how the introduction of the D3 will affect Canon's price. The sensor cost is mainly related to sensor area; you should be able to cramp more pixels into the same area without increasing the cost by much. As a studio camera, you only need 4 to 5 frames/second so that you don't have to have a lot of fast electronic to move those 20MP.

[ShunCheung]

Thom, thanks for your post and speak for yourself here. I probably should have said that you are "unhappy" or "disappointed" that you weren't invited. Hopefully Nikon USA will correct that in the future.

 

As far as in-body image stabilization goes, I think it would be difficult to achieve that on the FX sensor because the image circle is unlikely to be big enough in a lot of situations. The cameras that use in-body IS such as Minolta/Sony and Pentax are all DX type so far. Perhaps that is why Canon and Nikon are dismissing it at this point.

[mark_chappell]

<I>an action photographer would use the D300 for telephoto work where reach is critical,

and the D3 for low light work or where you already have a long enough lens in the 35mm

format.</I><P>

 

Fine, if you want to sacrifice bit depth or frame rate with the tele work (although I can't

believe that the D300 drops as low as 2.5 fps if you want 14 bits), or you are willing to

have two cameras handy if you want to switch subjects. My personal problem is that I

like to be able to shoot both landscape and wildlife on a given excursion, and walking

around with a heavy telephoto is hard enough without having to carry the additional

weight and bulk of an extra camera. For me, the 1.3X format is a useful compromise.

Others will see it differently.

[ShunCheung]

Mark, just like Thom, I too prefer a DX sensor, 8+ frames/sec DSLR with excellent AF for wildlife work. Bit depth being 12 should be just fine. The D300 may be the answer, but we'll see.

 

The DX sensor should be fine for nature landscape work, which is typically shot on a tripod. I have seen a lot of landscape work by John Shaw using the D2X, pritned into large panorams, and they look fine. The real problem for the DX sensor is low-light work: weddings, night or indoor sports .... The D3 seems to address that issue.

 

Otherwise, if you are that serious about landscape, maybe you should still be shooting 4x5 Velvia.

[tonybeach]

For Edward:

 

"14-bit files are no larger than 12-bit files. Any digital integer between 8 bits and 16 bits is represented in a 16-bit word."

 

I suspect you are ignoring compression algorithms. Seems to me with truly uncompressed data the files would end up being 60+ MB a piece.

 

For Mark:

 

"...a high FPS DX format 'pro' machine (D2X on steroids) with improved noise characteristics -- probably not feasible with small photosites -- would be even better."

 

Better noise characteristics than what? Canon has demonstrated that smaller photosites can deliver fairly impressive results. It remains to be seen, but I bet the D300 matches Canon's 30D, and that would be about a two stop improvement over the D2x.

 

"Fine, if you want to sacrifice bit depth or frame rate with the tele work..."

 

It remains to be seen what the difference will be between 12 bit and 14 bit measurements of the photosites coming from the same sensor will be. The D3 and the D300 will be the first cameras I'm aware of that will allow us to do that. Given that we regularly debate the merits of uncompressed NEF versus compressed NEF here, and the difference is also about 2 bits (the savings all concentrated in the highlights), I am skeptical that you will be able to demonstrate the difference.

[mark_chappell]

Like Shun, I'm not concerned that much with 12- vs. 14 bits, unless it's going to make some

obvious difference in image quality -- and that remains to be seen.

 

In terms of photosite dimensions and noise: I'm sure there are differences between C and N

in noise per unit photosite area -- or at least there have been in the past. But with a given

technology, you're still going to have an inverse relationship between noise and photosite

area. So Nikon probably could have improved the D2X noise levels with a DX sensor size, but

not as much as was possible by doubling the area of each photosite -- which they did with

the D3.

[Richard Williams]

Did any of you well-connected commentators get any impression of whether Nikon intends to make the FX sensor available further down the food chain in the medium term? (e.g. at the D300 level to compete with the 5D). Thom's comment about the naming of the grip seems to suggest that we might expect another camera in a similar 'form factor', though my uninformed guess would be that FX will be reserved for the single-digit models in this generation of cameras.

[ShunCheung]

<I>Did any of you well-connected commentators get any impression of whether Nikon intends to make the FX sensor available further down the food chain in the medium term?</I>

<P>

Richard, it seems to be logical that Nikon will eventually have "its own version of the 5D," i.e., an "economy model" of its FX-sensor DSLR. But then, once again, as Ellis points out above, those who really know the details are under NDA and won't post those details or would get into trouble. Those who don't know can speculate all day long but won't give you any really useful answer. It could happen in 2008, 2009 ....; it could be $2000 or $4000 .... Most of us simply don't know and those who know cannot tell.

[mawz]

@Edward Ingold: The last camera to pack 12bit values into a 16-bit word was the Pentax *istD, and that led to 13MB RAW files with a 6MP body(the *ist series also lacked compressed RAW files). All current RAW formats used a packed pixel method that simply stacks the pixels sequentially without padding them to 16 bits. This leads to significantly smaller RAW files for 12 bit output, and somewhat smaller files for 14 bit output.

 

Multiple compression options also don't mean uncompressed output is also an option. With todays DSP speeds, there's no good reason to chose uncompressed RAW over lossless compression, it costs you card space and write speeds with absolutely no gain. The real choice is between lossless compression (best quality, larger files, slower writes, slower processing) and lossy compression (lower quality, more speed and less space).

[gan_esh]

The is the first I'm hearing of 2.5 fps in 14bit mode for the D300.

 

I can imagine a Nikon PR guy saying, "It shoots at an amazing 6 fps BUT only 2.5 fps at it's highest quality setting." C'mon.... What the F*#K, man!

 

I hope it's not the same case with the D3. If Nikon suddenly say 9/11 fps in 12bit but only 5/6 fps in 14bit, I won't believe a single thing they say after that.

[Richard Williams]

'Those who don't know can speculate all day long but won't give you any really useful answer.'

 

Though Thom seems to be in the rather useful position of knowing a bit more than the rest of us from his sources without being bound by an NDA...

 

Regarding the 14-bit performance hit, is this more a consequence of the processor having to work internally on higher bit depth data, rather than the size of the output file? Has anything been said about whether the 14-bit framerate also drops in the D3?