Cost of DSLR components (info for us all)

[Progress Lover]

But it doesn't diminish the potential usefulness of allowing users generally to have a better understanding of what they're paying for and why.

We are talking about photography and a lot of useful info has come from this thread already even though it's not yet made the progress I would have liked.

Between us we have the knowledge to cost a DSLR and make a pretty good guess at it. Obviously there will be people on this forum that know the answers because they work or worked for a manufacturer. They may or may not be will to share the info and we can only speculate as to why.

I am wondering if Thom Hogan from the Nikon side would be able to help out with this. He has some pretty good knowledge of the subject and could surely at least illuminate the unit costs of the user manual etc. I'm sure he'd have plenty of other useful info too.

I guess it would be like walking into the lions den as too many around here would dismiss anything he said just because he's a Nikonian. However that hasn't stopped him before so let's hope he see's the thread.

I think that if/when this thread maxes out without reaching its aim, I will recap it and start again taking into account some of the points made on this thread.

With persistence it's just a matter of time (and wading through those claiming it's pointless) until it's done. Once it's done it's done and who knows how useful (or not) it will prove to be.

At the very least it will allow discussions to be based on a better understanding of the factors involved with the production of a DSLR and their relative costs.

Here's the list for anyone wishing to get back on topic:

LSI chip $15
Body casing + chassis + lens mount$??
LCD screen $??
Buttons $1?
Shutter unit (inc motor & mech) $35
Sensor(inc AA filter) $??
Mirror unit (inc motor & mech)$??
Prism/VF $??
Batteries (both main and button cell) $5?
AF system $??
Connectors (terminals on body) $??
Built in flash $??
Memory $??
A couple of circuit boards $1?
Fixings (screws etc) $1?
Putting it all together $??
Packaging $??
User manual $??
Cables/charger $5?
Bundled software $??
International logistics costs/unit (assume 300k units)$??

I'm sure that between us we have the knowledge to nail this.

If you can add a cost and a reason for your assertion, that would be great. Thanks.

As long as Cannon owns its patents, lits mfg. know how and its own
stock, this exercise is a waste of time . They can charge what
they want
for their product and I can say no thanks if I dont like the
product or its
price. They dont have to justify their price and I dont have to say
why
I do or dont buy their product----lets talk about photography.
--
Dmac

--
Keep photography wild.

 

[Progress Lover]

Probably nothing more than guesses which are worthless IMO.

Well then can you point us towards someone/something that could help? Or perhaps an explanation of your guesses might allow someone else to back them up or put them right - thus moving things along.

With an explanation, I can do more research and possibly find something that might verify or something else that sounds better. We might both learn something whilst increasing general knowledge in the process. A win win win.

--
Keep photography wild.

 

[Elwood_98034]

Nice one.
I suspect you'd need to add a little more for the hookers in this
case. ;-)

Maybe I should revise the list to include corporate hookers!

I still feel that whilst this type of method is known to work, it
is a little hypothetical for the average joe.

My objective is to get into the public domain and decent breakdown
of costs in order that the average joe can better conceptualise
exactly where all their money goes.

I feel this could offer a major advance in general understanding of
the real costs involved in these products and therefore why it can
cost very little indeed to double performance (as the component
costs are a tiny fraction of the cost to the end user after all is
said and done and therefore doubling the component costs (or more)
only needs to make a small difference in the shelf price).

After all, it doesn't cost twice as much to assemble better
components or to market or distribute them and the same hookers
will probably be employed for the 1Ds team as the Xti team etc etc.
He he.

No, the hookers used at a management level are far finer than those used by the workers, and the workers are usually expected to pay their own way in that regard, or even seek it out independently. Corporate entertainment at a management level is usually supplied as a perk, and is something that the average worker never even hears about, let alone gets to partake in. A lot of 'entertainment' is provided on a discreet basis at 'management get-aways', yacht trips, things like that, where secretaries and lower level employees don't get to attend.

A far greater amount of money is spent on management perks like BMW's, Tag Heuer watches as bonus type incentives (non-taxable as a corporate gift, and no records kept), first class airplane tickets, sales conferences in places like the Swiss Alps and Tahiti, clothing allowances, subsidised housing, spouse air fares, company credit cards (with say a $2,000/month limit, and no receipts required), vouchers to Disneyland, discounts on car purchases, you name it.

Keep photography wild.

--
Elwood.

Light! Give me light!

 

[David Hull]

If I think of anything, I'll throw it in. Like the other guy in this subsection, I have generally found this to be a waste of time unless it is your job to do it and you have the backing of some pretty sophisticated parts databases. As someone else pointed out, I think you would be better off (and closer in the long run) with one of the "rules of thumb". The trouble is, I don't really know what applies to something like a camera. I quoted teh 5X rule taht some use, but a camera is not like a Satellite box or a cable modem, the camera has a lot of little tiny screws and stuff and requires skilled assembly -- must cost more IMO.

 

[David Hull]

I don't think what you are doing does any harm at all, I do, however, based on past experience with this sort of thing wonder how accurate it will be when it ends up. I have often had to do one of these on something relitively simple like a TV tuner -- count the parts, put into Excel and add up the cost. A camera is a lot more complex, requiring a lot of guessing which reduces the accuracy.

What would be more interesting to me is a pie chart showing where the largest costs are. Is it really the sensor, or is it all of that mechanical crapola (magnesium body parts etc.)

Another thing that would be intersting is, how different are the cameras really. For example, the shutter in the 30D is rated at 100k clicks. Is it different than the one in my 20D or is the same one and they just fessed up to the number.

 

[ceoyoyo]

Chips become cheap after you've made millions of them, because you get to divide the price of the fab and design over each chip.

The Digic chip will get somewhat cheap because Canon uses it in everything, but not cheap like a processor for a computer. Intel sells millions of Core chips each year. How many cameras does Canon sell, total, in a year? The sensor will be less cheap, because it's bigger, more prone to failure and only used in a specific line of cameras.

Chips are not cheap. Ones that sell a LOT are cheap. Others can be VERY expensive. I was putting together a card once and needed an obscure chip that was on the order of $10,000 for ONE.

Thank you for illustrating my point about assumptions and guesses based on, well, no knowledge.

--
Robb

 

[Progress Lover]

I don't think what you are doing does any harm at all, I do,
however, based on past experience with this sort of thing wonder
how accurate it will be when it ends up. I have often had to do
one of these on something relitively simple like a TV tuner --
count the parts, put into Excel and add up the cost. A camera is a
lot more complex, requiring a lot of guessing which reduces the
accuracy.

What would be more interesting to me is a pie chart showing where
the largest costs are. Is it really the sensor, or is it all of
that mechanical crapola (magnesium body parts etc.)

Another thing that would be intersting is, how different are the
cameras really. For example, the shutter in the 30D is rated at
100k clicks. Is it different than the one in my 20D or is the same
one and they just fessed up to the number.

I'm sure that accross all industries there are plenty of marketing tactics that are though of after the fact and/or held back as ammo for later. They get added when an opportunity or neccessity arises.

The 20D had enough going for it that it was clearly a great advance at the price point. Baring that in mind, why boast about the shutter life when you could hold back that boast for the next 'upgrade'?..

In a way it stinks. In a way it's common sense and 'good business' (ironic that the word good is used to describe something that is verging on the opposite). Just depends which side of the fence you're on.

In my years of marketing financial services I would always attempt to clinch a deal whilst keeping as many tricks as possibly l up my sleeve just in case.

Using them all unnecessarily would mean I couldn't use them later if the need for something more arose. This is absolutely basic stuff and it would be crazy to think that all major corporations don't do it (sadly - if you're a customer).

Regarding pie charts. I'll produce one if I can get enough info to get reasonable estimates of the various costs.

It is almost a certainty that between the more experienced users on this forum we can handle a task even of this complexity if we put our minds to it.

--
Keep photography wild.

 

[Vegasluvr]

That is like listing all parts involved to build a vehicle. People forget that in the price of a DSLR, you are also paying for the company's research amd development costs, production costs, raw materials costs, employee wages, overhead costs and so forth.

On a digital camera, the primary and most expensive parts are your PCB assemblies, the CCD/CMOS sensor, body (the frame), Main CPU, drive unit, flash assembly if it has a flash built in and then the lens mount.

When we damage these parts, we have to take into consideration the cost for Canon to repair the unit (parts and labor).

 

[Progress Lover]

I don't know if it must cost more. I guess it does depend on the fab techniques and where they are. Someone will know this.

My guess is that they're put together in China. Probably in something very much like a scaled down car plant.

Basic components are probably assembled into the building blocks of the camera automatically in japan then shipped to the factory that assembles these blocks into finished bodies.

A human probably grabs a block, sticks it (with the ease of having done it 1000x/day - that's twice a minute) into the body and tightens literally a couple of screws with a power-tool and maybe plus a std connector into a std socket somewhere. Then it's on to the next guy who does the same with the next building bock.

Whilst this type of work is indeed skilled insomuch as it can't be done without training/practice, it's highly likely that each fabrication technician literally only has to know how to slot one thing into a space and tighten a couple of screws and maybe plug in the same same mini-connector into the same socket every time.

Under those circumstances in a far eastern factory, the price for assembly would be absolutely negligable on a per unit basis.

If 100 guys can build 1000 cams/day then you've got $1000000 worth of product built (put together - not inc component costs r&d etc) for $1000!

I can strip a camera down and rebuild it in pretty short order with no training, but just a little experience of tinkering with all sorts of hotographic kit over the years. If I did it all day every day and only had to deal with one particular component (attaching battery doors or pop-up flashes etc), then I would be seriously good at it in no time and with all the correct tools (instead of my usual improvisations), it would be a cinch.

I think that the upshot of this excercise when completed will be that it will be quite surprising to most users how little it costs per unti to design and build a DSLR. The type of costs that make up the balance and the proportion of the retail price they comprise will be suprisingly high to most people.

Knowing this has certain implications for people's expectations. Obviously if they understand the costs involved then they will know that adding a feature doesn't necessarily need to change the cost of the unit on the shelf at all.

Often people claim that progress in DSLR design can't be made because it would be too expensive. This is clearly not the case with a lot of things which are either firmware related or require changing the current component combo to a new combo that costs literally only a few $ more.

Obviously as tech advances it can actually be cheaper to make better cameras (reflected by reduction in price but increase in performance to date). If a new more powerful chip is cheaper than the old chip then using the new chip in the next camera will reduce costs AND increase performance.

This is true in the computer industry in general and DSLRs and now largely computers.

The price of computer components is more dynamic than the price of DSLRs. The price of a given DSLR model does not drop as the price of its components drop. It drops in response to competition. This means that it often makes sense to allow older cameras to dwell on the market as the component costs have gone halved, but you're still selling the thing for 80% of the release price... - I won't mention any particular models here, but I'm sure there are a couple of pretty obvious examples in the industry that fit the bill.

This situation is a disincentive to progress as it means that new models can be (and are) held back to profitably milk the existing ones.

If component prices were transparent, then prices of DSLRs could be more dynamic in their reflection of component prices (just as computer prices dynamically move relative to component prices).

This would make it non-sensical to delay production and release of new products and that would be good for progress and for photographers.

Costing a DSLR and getting that basic knowledge into the public domain is a good first step to better knowledge and transparency of component costs. That in turn can make DSLR prices more dynamic and reflective of those costs and that can help to drive progress in the industry by disincentivising the activity of holding back products from a financial point of view (although obviously it will often still make strategic sense for the same reasons we mentioned at the top of this post - keeping stuff up one's sleeve).

If I think of anything, I'll throw it in. Like the other guy in
this subsection, I have generally found this to be a waste of time
unless it is your job to do it and you have the backing of some
pretty sophisticated parts databases. As someone else pointed out,
I think you would be better off (and closer in the long run) with
one of the "rules of thumb". The trouble is, I don't really know
what applies to something like a camera. I quoted teh 5X rule taht
some use, but a camera is not like a Satellite box or a cable
modem, the camera has a lot of little tiny screws and stuff and
requires skilled assembly -- must cost more IMO.

--
Keep photography wild.

 

[Progress Lover]

No I didn't forget. And yes it's exactly like listing the cost of parts to build a vehicle.

The reason is that if you know that, then quid-pro-quo you know the other non-component costs that make up the balance. The point is that this would be a total shock to most people.

The benefit is that it alters perception. People will no longer believe that a given feature can't be added because it would be too expensive.

If then know that a 100k click shutter costs $30 and a $300k shutter costs $40 then they're not going to suggest that it would be "impossible" to put a longer life shutter in a cam without it making the cost prohibitive...

Likewise if they know that a basic AF unit costs $10 and the one in the 1D costs $20 they are going to want to know why they can't have that feature for under $4.5k...

You have to remember that r&d costs on most of the mechanical parts were done years ago and are now almost nil. Also you have to average them accross all the units produced. So, even if you spent $50m developing a shutter (figure entirely hypotheical and for the sake of an example - probably far too large), when you produce millions of them (over many years) that's only a few bucks each...

People are confused by the big numbers used to define r&d versus how little that affects the unit cost of a product. As a result they often assume that things cost far more than they do especially in light of the retail price of what they bought.

This is an attempt to throw some light on the situation and further all our understanding of which factors really impact the cost of a product and therefore what is reasonable and unreasonable to expect.

Potentially a useful exercise for someone interested in the photo business.

That is like listing all parts involved to build a vehicle. People
forget that in the price of a DSLR, you are also paying for the
company's research amd development costs, production costs, raw
materials costs, employee wages, overhead costs and so forth.

On a digital camera, the primary and most expensive parts are your
PCB assemblies, the CCD/CMOS sensor, body (the frame), Main CPU,
drive unit, flash assembly if it has a flash built in and then the
lens mount.

When we damage these parts, we have to take into consideration the
cost for Canon to repair the unit (parts and labor).

--
Keep photography wild.

 

[Progress Lover]

Chips become cheap after you've made millions of them, because you
get to divide the price of the fab and design over each chip.

Absolutely. Just what I was getting at. Thanks.

And whilst Canon don't make millions, they make a lot. Evidently Fujitsu can make LSI chips for $20 or less. I'm sure Canon are competitive and maybe even have an advantage in selling more cameras than the companies Fujitsu is aiming its chip at. After all it's for 3 layer sensored DSLRs to the volume isn't going to be as large as Canon's...

Could well be that Canon's LSI chip costs $10. Have you any evidence to help work it out either way?

The Digic chip will get somewhat cheap because Canon uses it in
everything, but not cheap like a processor for a computer.

You assume (incorrectly) that it's as powerful and advanced as the 'processor for a computer' (it is a 'processor for a computer' BTW but I get your drift meaning desktops etc).

Intel
sells millions of Core chips each year. How many cameras does
Canon sell, total, in a year? The sensor will be less cheap,
because it's bigger, more prone to failure and only used in a
specific line of cameras.

Chips are not cheap. Ones that sell a LOT are cheap. Others can
be VERY expensive. I was putting together a card once and needed
an obscure chip that was on the order of $10,000 for ONE.

Well I don't think anyone is suggesting that any chip in the 30D costs $10k now are they which begs the question:

However much of a truism it may be to state that it is possible for a chip to be expensive, how relevant is your comment to the discussion at hand?

Clearly none of the chips in the 30D costs over $40 and in reality it's probably closer to $15.

Do you have any reliable information as to the cost of any component of the 30D or any other camera/modern electronic product that could be considered a good parallel for the purposes of this discussion?

--
Keep photography wild.

 

[Viking9]

--
==============================
Tom Drake
Suffolk UK
'The man who has never made a mistake has never made anything'

 

[Progress Lover]

Youch. Thinking about it what you say could be true.

However even with this level of corporate indulgence on a wide scale, the cost would only amout to a tiny fraction of the unit cost of a product.

I'll include $5/camera for hookers and perks just so no-one can say we didn't take it into account! Lol.

If anyone has official stats on hooker usage and total perk cost at Canon and can give me a more accurate figure than with good evidence, then we'll look at it again!

BTW $5/unit would give Canon's photographic industry executives an annual 'hooker budget' of about $70m between them ($5 per cam/lens - inc p&s digicams).

http://news.com.com/Digicam+sales+outshoot+expectations/2100-1041_3-6145329.html

Canon had about 1/5th of 30m digicam sales last year in the US alone. Just $5 on each of those would amount to $30m for hookers etc. Then include the rest of the world...

Canon's head of corporate pimping has quite a budget.

Anyway, here's the revised list:

LSI chip $15
Body casing + chassis + lens mount$??
LCD screen $??
Buttons $1?
Shutter unit (inc motor & mech) $40
Sensor(inc AA filter) $30
Mirror unit (inc motor & mech)$??
Prism/VF $??
Batteries (both main and button cell) $5?
AF system $??
Connectors (terminals on body) $??
Built in flash $??
Memory $??
A couple of circuit boards $1?
Fixings (screws etc) $1?
Putting it all together $??
Packaging $??
User manual $??
Cables/charger $5?
Bundled software $??
International logistics costs/unit (assume 300k units)$??
Corporate perks/ retrieving Tag Heurs from Swiss hookers on yachts $5?

I'm sure that between us we have the knowledge to nail this.

NB those hookers are effectively stealing a backup battery from each and every one of us!

Anyway with a little more seriousness, can anyone throw any more light on the likely cost of any given component and why they believe their info to be accurate? Thanks.

Nice one.
I suspect you'd need to add a little more for the hookers in this
case. ;-)

Maybe I should revise the list to include corporate hookers!

I still feel that whilst this type of method is known to work, it
is a little hypothetical for the average joe.

My objective is to get into the public domain and decent breakdown
of costs in order that the average joe can better conceptualise
exactly where all their money goes.

I feel this could offer a major advance in general understanding of
the real costs involved in these products and therefore why it can
cost very little indeed to double performance (as the component
costs are a tiny fraction of the cost to the end user after all is
said and done and therefore doubling the component costs (or more)
only needs to make a small difference in the shelf price).

After all, it doesn't cost twice as much to assemble better
components or to market or distribute them and the same hookers
will probably be employed for the 1Ds team as the Xti team etc etc.
He he.

No, the hookers used at a management level are far finer than those
used by the workers, and the workers are usually expected to pay
their own way in that regard, or even seek it out independently.
Corporate entertainment at a management level is usually supplied
as a perk, and is something that the average worker never even
hears about, let alone gets to partake in. A lot of 'entertainment'
is provided on a discreet basis at 'management get-aways', yacht
trips, things like that, where secretaries and lower level
employees don't get to attend.

A far greater amount of money is spent on management perks like
BMW's, Tag Heuer watches as bonus type incentives (non-taxable as a
corporate gift, and no records kept), first class airplane tickets,
sales conferences in places like the Swiss Alps and Tahiti,
clothing allowances, subsidised housing, spouse air fares, company
credit cards (with say a $2,000/month limit, and no receipts
required), vouchers to Disneyland, discounts on car purchases, you
name it.

Keep photography wild.

--
Elwood.

Light! Give me light!

--
Keep photography wild.

 

[Progress Lover]

I'll tell you whether you'd have been better off on the forums.

What are you shooting in Suffolk anyway? Tractors? Grey skies, brown water and reeds? Murdered hookers (topical in that area in case anyone's wondering)? Or are has the pace picked up there recently? Maybe it's time for the farmer's show?! YEAH!

Anyway, show us what you came up with on your outing in beautiful Suffolk. Nothing like a nice plowed field to get the heart racing. Lol.

--
==============================
Tom Drake
Suffolk UK
'The man who has never made a mistake has never made anything'

--
Keep photography wild.

 

[duncan74]

I'll tell you whether you'd have been better off on the forums.

What are you shooting in Suffolk anyway? Tractors? Grey skies,
brown water and reeds? Murdered hookers (topical in that area in
case anyone's wondering)? Or are has the pace picked up there
recently? Maybe it's time for the farmer's show?! YEAH!

Anyway, show us what you came up with on your outing in beautiful
Suffolk. Nothing like a nice plowed field to get the heart racing.
Lol.

Progress Lover,

Sorry, I can't add any valid information to your quest, but please can you make sure you carry on with this. It's bloody hilarious and you're sharp wit and incisive social comment are so entertaining.

Oh, and I'll help you with a bit of a set up for your next post, I live in Manchester, haven't sold any images (but get enjoyment and relaxation out of just taking photos) and support Hull City 'soccer' team.

(Oh,and have you included the cost of the instruction book, and translators?)

 

[Steve Balcombe]

What are you shooting in Suffolk anyway? Tractors? Grey skies,
brown water and reeds? Murdered hookers (topical in that area in
case anyone's wondering)?

You have crossed the line, Progress Lover.

 

[MichaelOuellet]

Your point is interesting but it depends if you have all these gears to make money or for your leisure, or both...

I'm not a professional photographer but some of my projects contribute to pay my equipment.

If I compare myself to my friends spending their money on Cars or at a Casino for their leisure, photography can be a cheap recreation.

A friend of mine asked to meet me to learn about his camera and lens because he needs some explainations. He has a 5000$ investment and used it once on a 2 weeks travel for the last 2 years.

So my point is that for this person, his camera has a different "value" then me, even with a high cost.

This confirms that trying to find the "cost" of a camera is futile for me.
Value is more important.

--
Michael Ouellet
Quebec city, Canada
http://michael.volcan.ca/

 

[thecameraguy]

Clearly none of the chips in the 30D costs over $40 and in reality
it's probably closer to $15.

On what basis do you arrogantly assume you KNOW that "none of the chips in the 30D costs over $40?.. it would surprise me not at all for the sensor to be $80-$100..

--
~ Being over-exposed can get you arrested ~

 

[thecameraguy]

Dude,

When you are embarked upon some totally futile reverse engineering cost analyis based upon zero knowledge, little information and no business understanding you really should expect people to point out the futility of the exercise.

I am glad to see you care more about your petty project than the families of the murder victims.

You are a sad individual

What are you shooting in Suffolk anyway? Tractors? Grey skies,
brown water and reeds? Murdered hookers (topical in that area in
case anyone's wondering)? Or are has the pace picked up there
recently? Maybe it's time for the farmer's show?! YEAH!

Anyway, show us what you came up with on your outing in beautiful
Suffolk. Nothing like a nice plowed field to get the heart racing.
Lol.

--
==============================
Tom Drake
Suffolk UK
'The man who has never made a mistake has never made anything'

--
Keep photography wild.

--
~ Being over-exposed can get you arrested ~

 

[pavi1]

Gives you a good starting point.
--



Everything happens for a reason. #1 reason: poor planning